Decoding method and decoding system based on weight adjustment algorithm parameter

Abstract

The present disclosure provides a decoding method and a decoding system based on weight adjustment algorithm parameters. The method is applied to a decoder to generate an MxN low-density parity-check code from an input signal, where there are N variable nodes and M check nodes. In the method, the information of the variable nodes and the check nodes is initialized. Through multiple iterations, messages from each variable node to multiple check nodes are formed. After excluding the to-be-calculated connections, the remaining connections are calculated and summed. The multiple variable nodes are updated. Through multiple iterations, messages from each check node to multiple variable nodes are formed. After excluding the to-be-calculated connections, the remaining connections are calculated and multiplied. According to the multi-dimensional parameters, an estimated first minimum value or an estimated second minimum value is obtained to update the messages from the check nodes to the variable nodes, and a decision is made.

Description

TECHNICAL FIELD

[0001] The present disclosure provides a decoding technique, in particular, a decoding method and a decoding system for modifying algorithm parameters in a decoder to improve decoding performance. BACKGROUND

[0002] Low density parity check (LDPC) codes are error correction codes that can approach the Shannon limit, which is the maximum transmission rate under a given noise level, and thus have become the most popular error correction codes. LDPC codes can be found in various systems that require decoding, such as IEEE 802.11n (wireless local area network), satellite television systems, and IEEE 802.3an (10 Gbps Ethernet communication over unshielded twisted pair), etc.

[0003] The best decoding performance of LDPC codes is achieved by using a belief propagation (BP) iterative soft decoding, which is a sum-product (SP) algorithm. However, the hardware complexity of the conventional sum-product algorithm is too high, and thus a simplified version, such as a min-sum (MS) algorithm, has been developed. However, although the hardware complexity of the min-sum algorithm is greatly reduced, it also causes a serious performance degradation problem, and thus a normalized min-sum (NMS) algorithm and an offset min-sum (OMS) algorithm have been developed to improve the performance degradation problem, so that only a small amount of hardware complexity is added, but the performance is as good as the sum-product algorithm, and thus the normalized min-sum and offset min-sum algorithms have become the most popular algorithms.

[0004] However, although the normalized min-sum algorithm and the compensated min-sum algorithm can provide a lower complexity decoder architecture, the min-sum algorithm requires searching for the first minimum value (min1) and the second minimum value (min2) when calculating the check node update, and the complexity of this operation is completely determined by the check node degree, i.e., the number of variable nodes covered by the check equation.

[0005] For example, in a 10 Gbps Ethernet system, the check node degree in the LDPC decoder architecture is 32, i.e., the first minimum value (min1) and the second minimum value (min2) must be found from these 32 variable nodes each time the check node update is performed, and this operation limits the clock rate, latency, hardware area, iteration number, and performance of the decoder.

[0006] The first minimum value (min1) is the minimum value, and the second minimum value (min2) is the second smallest value, where searching for the minimum value is relatively small, and searching for the second smallest value requires a sorting process, so the processor has a large burden, and the more the number of variable nodes, the greater the amount of calculation. In order to avoid the burden of searching for the second minimum value, the prior art further develops a single-min algorithm (SMA), and the single-min algorithm architecture mainly modifies the behavior of the check node update in the min-sum algorithm. The single-min algorithm no longer searches for the second minimum value, but only searches for the first minimum value, and instead estimates the second minimum value using the single-min algorithm, in other words, after estimating the second minimum value, it is called the estimated second minimum value (which can be referred to as min2 est ), which is equivalent to a scrambled second minimum value.

[0007] When the estimated second minimum value (min2 est ) is properly generated, it has the effect of reducing the error floor of the LDPC code. The error floor phenomenon is that when the error rate of the LDPC code is low enough (for example, in IEEE 802.3an, the error rate usually occurs at BER = 10 -10 , FER = 10-8 When the error rate reaches a certain level, the error floor phenomenon will begin to occur. The error floor phenomenon is caused by the trapping set or the absorbing set. The error floor phenomenon is not good for the performance of the system. The single minimum algorithm can slow down the error floor phenomenon by adding noise to increase the chance of escaping the trapping set. SUMMARY

[0008] The disclosure provides a decoding method and a decoding system based on weight adjustment algorithm parameters. The decoding method and the decoding system have the advantages of the single minimum algorithm with improved decoding performance. The decoding method and the decoding system modify the two-dimensional variables of the single minimum algorithm to three dimensions by weight adjustment. The decoding method and the decoding system provide a modified single minimum algorithm. The decoding method and the decoding system increase the range of use, support more modulation methods, and obtain a wider range of fixed points.

[0009] According to an embodiment, the decoding method based on weight adjustment algorithm parameters is applied to a decoder. An input signal generates an M×N low-density parity-check code (LDPC) with a plurality of (N) variable nodes and a plurality of (M) check nodes. The method includes initializing the information of the plurality of variable nodes and the plurality of check nodes. After a plurality of iterations, the information of each variable node to the plurality of check nodes is formed. After excluding the connections to be calculated, the remaining connections are calculated and summed to update each variable node according to the information of the plurality of check nodes connected thereto. Then, after a plurality of iterations, the information of each check node to the plurality of variable nodes is formed. After excluding the connections to be calculated, the remaining connections are calculated and multiplied to update each check node according to the information of the plurality of variable nodes connected thereto. Then, the inner product is calculated according to the estimated first minimum value or the estimated second minimum value. The information of the check node to the variable node is obtained, and a decision is made based on the information.

[0010] In the method of obtaining the estimated first minimum value and the estimated second minimum value, the minimum value of the updated plurality of variable nodes is searched to obtain the first minimum value. The first minimum value is used to obtain the pseudo second minimum value. The first parameter (α) is multiplied by the obtained first minimum value to obtain the estimated first minimum value. The second parameter (β) is multiplied by the first minimum value, and the third parameter (γ) is multiplied by the pseudo second minimum value to obtain the estimated second minimum value.

[0011] The first parameter (α), the second parameter (β), and the third parameter (γ) satisfy the relationship (β+γ)≥α. The equation applied is as follows, where 'N' is the number of variable nodes; 'M' is the number of check nodes; 'n' is the variable node number; and'm' is the check node number. information for variable node n to check node m; 'n''represents the remaining variable node number excluding the connection to be calculated; information for check node m to variable node 'n''after excluding the connection to be calculated, that is, v2c information; the sign function'sign()' returns 0, 1 or -1 according to the value in the function being 0, positive or negative;'min1' is the first minimum value; the function is a function for taking the minimum value;'min1 est is the estimated first minimum value;'min2 est is the estimated second minimum value;'min2"'is the pseudo second minimum value.

[0012] For m∈{1,…M}and n∈N m ;

[0013]

[0014]

[0015]

[0016] min1 est = α · min1;

[0017] min2 est = β · min1 + γ · min2" '.

[0018] Preferably, the method for judging the estimated first minimum value or the estimated second minimum value is determined according to the judgment result of whether the variable node number in the information of the check node to the variable node is the smallest variable node number in the information position of the check node.

[0019] And, the information of the variable node can be obtained by adding the intrinsic information of the variable node and the information of the check node to multiple variable nodes updated by the connection of the check node to other variable nodes, and then making a decision.

[0020] For further understanding of the features and technical contents of the present application, please refer to the following detailed description and drawings of the present application. However, the drawings provided are only for reference and illustration, and are not used to limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 A circuit architecture embodiment diagram of a decoding system is shown;

[0022] Figure 2 An example schematic diagram of a Tanner graph showing the decoding process of a low-density parity-check code is shown;

[0023] Figure 3 An example Tanner graph illustrating the sum operation in the decoding process of a low-density parity-check code is shown;

[0024] Figure 4 An example Tanner graph illustrating the product operation in the decoding process of a low-density parity-check code is shown;

[0025] Figures 5 to 8 A logic circuit block diagram illustrating the calculation of the first minimum value and the pseudo second minimum value is shown;

[0026] Figure 9 A logic circuit block diagram illustrating the implementation of the modified single minimum and minimum sum algorithm is shown; and

[0027] Figure 10 A flowchart illustrating an embodiment of the decoding method based on weight-adjusted algorithm parameters is shown. DETAILED DESCRIPTION

[0028] The following describes embodiments of the present application by way of specific examples. Those skilled in the art can understand the advantages and effects of the present application from the disclosure herein. The present application can be implemented or applied by other different embodiments, and the details in the description can be modified and changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. In addition, the drawings of the present application are merely simple schematic illustrations and are not drawn to scale. The following embodiments will further describe the relevant technical content of the present application in detail, but the disclosed content is not intended to limit the scope of protection of the present application.

[0029] It should be understood that although the terms "first", "second", "third", etc. can be used herein to describe various elements or signals, these elements or signals should not be limited by these terms. These terms are mainly used to distinguish one element from another element or one signal from another signal. In addition, the term "or" used herein can include any combination of one or more associated listed items as appropriate.

[0030] The present disclosure discloses a decoding method and a decoding system based on weight-adjusted algorithm parameters, wherein a modified single minimum and minimum sum algorithm (SMAMSA) is proposed. By using more dimensions of variables than conventional algorithms, the proposed algorithm can improve performance, increase the range of application, achieve lower error rates, reduce hardware complexity, and reduce power consumption.

[0031] The decoding system architecture diagram for implementing the decoding method based on weight-adjusted algorithm parameters can refer to Figure 1As shown, in the process of signal transmission of the applied communication system, in order to check the possibility of data destruction due to interference of the transmission medium and thus reduce the reliability of signal transmission, error correction code (error correction code) such as low density parity check code (LDPC) is used during transmission, that is, redundant information is added during signal transmission, and the receiving end infers the correct information from the received information to restore the destroyed information.

[0032] The decoding system includes a decoder at the receiving end, which receives signals such as communication signals through an input circuit 101, and after preliminary processing, outputs to an LLR (log-likelihood ratio) operator 103, which will provide a log-likelihood ratio, the purpose of which is to obtain a lower error rate and higher performance. In addition, the log-likelihood ratio (LLR) can be decoded and scaled to provide the low-density parity check code (LDPC) 105 with correct log-likelihood ratio, and in the decoding aspect, the connection relationship between the check nodes and variable nodes is updated and iterated multiple times, and finally the probability of the signal content is calculated, and finally the decoded signal is output through the output circuit 107.

[0033] In order to highlight the technical features of the decoding method and decoding system based on weight adjustment algorithm parameters, first describe the differences between the single minimum algorithm and the min-sum algorithm in the current technology, and the following example is the min-sum algorithm decoding process.

[0034] The algorithms disclosed in the book include the modified single minimum min-sum algorithm (SMAMSA) applied in the decoder, and in the min-sum algorithm, an MxN low-density parity check code is proposed, which has N variable nodes and M check nodes. It is pointed out here that the complexity of the operation is determined by the degree of the check node, that is, the number of variable nodes covered by the check equation. For related calculations, please refer to Figure 2 The example shows the Tanner graph used to demonstrate the decoding process of the low-density parity check code.

[0035] Figure 2The decoder shown in the figure is a decoder that applies the concept of message passing to the decoding process, in which probabilities are calculated by each of the variable nodes 21 and the check nodes 22 and then transmitted to the other. In the figure, the probability from a certain variable node to a certain check node (for example, from the nth variable node 211 to the mth check node 221) is determined by all the check nodes connected to the nth variable node 211 except the mth check node 221.

[0036] The equations for the decoding algorithm of the low-density parity-check code are described below. Let n be the variable node number, m be the check node number, N m N (m) represents all the variable nodes 21 participating in the current check equation m (the mth check equation); M n M (n) represents all the check equations (check nodes (M) 22) participating in the current variable node n (the nth variable node).

[0037] Figure 2 The plurality of connections between the plurality of variable nodes 21 and the plurality of check nodes 22 are schematically shown, and the decoding process of the low-density parity-check code is performed by calculating the probabilities by each of the variable nodes 21 and the check nodes 22 and then transmitting the information to the other. In the connection between the nth variable node 211 and the mth check node 221, for example, the intrinsic information is written in the plurality of variable nodes 21 in the initialization process, which is the log-likelihood ratio (LLR) of the decoder and is used to express that the value of the variable node is close to 0 or close to 1. i = k represents the kth iteration in the low-density parity-check code (LDPC) decoding process. v2c (n, m, k) represents the v2c information 201 from the variable node (the nth variable node 211) numbered n in the kth iteration to the check node (the mth check node 221) numbered m, in which v2c represents the abbreviation of variable node to check node. c2v (m, n, k) represents the c2v information 202 from the check node (the mth check node 221) numbered m in the kth iteration to the variable node (the nth variable node 211) numbered n, in which c2v represents the information from the check node to the variable node.I nrepresents the intrinsic information of variable node n, intrinsic information represents the original information when entering the system; a is a normalization factor, wherein, in a general min-sum algorithm (MS), a = 1; in a normalized min-sum algorithm (NMS), a ≠ 1.

[0038] The following are the calculation steps of each stage of the min-sum algorithm.

[0039] In the initialization stage (Initialization), the intrinsic information is written for each variable node, and for the check node, i = 0 at the beginning, and when the iteration is not performed (0th iteration), as equation one, = 0 (the check node is initially 0), wherein the symbol represents any one, and ∈ represents belonging to, represents belonging to N m any n.

[0040] Equation one:

[0041]

[0042] First step: variable-node update, as equation two, the variable-node update gives the information of the check node (v2c information), wherein there are N variable nodes, M check nodes, n is the variable node number, and m is the check node number.

[0043] Equation two:

[0044] For n ∈ {1, … N} and m ∈ M n ;

[0045]

[0046] Further, in the kth iteration, the information of the variable node to the check node (the mth) is formed, after excluding the to-be-calculated connection, the remaining connections are summed, that is, the remaining connections are the variable nodes participating in the mth check node (check equation). The schematic diagram can be referred to Figure 3 The Tanner graph is used to demonstrate the example schematic diagram of the sum of the decoding process of the low-density parity-check code. Taking the information transmitted by the nth variable node 211 to the mth check node 221 as an example, in the process of forming the information of the variable node to the check node by multiple iterations, that is, excluding the connection represented by the v2c information 201 in the figure, but obtaining the information (c2v information) formed by the connections (301, 302, and 303) of the other several check nodes 311, 312, and 313 to the nth variable node 211 and the intrinsic information (In ) and, to synthesize the updated variable node information

[0047] Second step: check-node update, as equation three, that is, update the mth check node 221 to the nth variable node 211 information (c2v information 201). The sign function in the equation is to return 0, 1 or -1 according to the value in the function is 0, positive or negative.

[0048] Equation three:

[0049] For m∈{1,…M}and n∈N m ;

[0050]

[0051] The kth iteration, Forming the check node to the variable node information, excluding the link to be calculated, calculating the product of the remaining link, finding the minimum value. The schematic diagram can refer to Figure 4 The Tanner graph is used to demonstrate the example of the product in the decoding process of the low-density parity-check code. Taking the transmission of the mth check node 221 to the nth variable node 211 as an example, in the process of forming the check node to the variable node information , excluding the c2v information 202 link in the figure, but getting the product of the information (v2c information) formed by the link (401, 402 and 403) of the other three variable nodes 411, 412 and 413 to the mth check node 221, and finding the minimum value, to get The information (c2v information) of the mth check node 221 to the nth variable node 211 is updated.

[0052] In the decision-making stage, all the information obtained above is added up to make the final decision. The first step uses the hard decision algorithm to decode, and each input and output signal is represented by 1 or 0. As equation four, the internal information (I n ) of the variable node and the information of the check node to the variable node updated by the link with other variable nodes are added up To get the variable node information (v n ) to make the final decision: D n Is 0 or 1.

[0053] Equation four:

[0054]

[0055] Third step: modify the single-min algorithm min-sum algorithm (SMAMSA) in the second step of the initialization phase above, in which the check node update is modified to equation five.

[0056] In equation five, the check node update, or updating the c2v information, where there are N variable nodes and M check nodes; n is the variable node number, and m is the check node number.

[0057] Equation five:

[0058] For m∈{1,…M}and n∈N m ;

[0059]

[0060]

[0061]

[0062] min1 est =α·min1;

[0063] min2 est =min1 est +γ·min2″′。

[0064] Where, represents the information from the check node to the variable node, and the min1 est or min2 est , when the variable node number n in is not the position of the minimum v2c information (the information from the variable node to the check node), min1 est is used, otherwise, min2 est is used, and then the inner product with the updated variable node is calculated to update the information of the check node where α and γ are operation parameters to obtain min1 and min2.

[0065] In the aforementioned single-min algorithm, the min2 (the second smallest value) is generated by using the pseudo-min2 (min2 est) to replace the second minimum value (min2) that would otherwise need to be searched for in the min-sum algorithm (MS).

[0066] The search for the first minimum value (minl) can be performed using any of the circuits shown in Figures 5 to 8 but does not exclude searching for the first minimum value in another way. In particular, the first minimum value (minl) can be obtained in any of the circuits shown in Figures 5 to 8 In addition to obtaining the first minimum value (minl) in any of the circuits shown in

[0067] According to the several embodiments of the block diagram of the logic circuit with a check node degree of 16 shown in Figures 5 to 8 Figure 5 shows that 16 input signals are input to 4 calculation units M41, the minimum values calculated by the respective calculation units are compared, and the first minimum value (minl) is obtained by inputting the minimum values calculated by the calculation units M42 to which the minimum values calculated by the calculation units M41 are input to the calculation units M42. However, instead of calculating the second minimum value (min2), the pseudo second minimum value (min2'") is obtained on the basis of the additional information obtained in the calculation process. Figure 6 shows that 16 input signals are input to 8 calculation units M21, the minimum values calculated by every 4 calculation units M21 are input to the next stage 2 calculation units M41, and the first minimum value (minl) and the pseudo second minimum value (min2'") are obtained by the next stage calculation units M22. Figure 7 shows that 16 input signals are input to 8 calculation units M21, the minimum values calculated by every 2 calculation units M21 are input to the next stage 4 calculation units M21, and the first minimum value (minl) and the pseudo second minimum value (min2'") are obtained by the next stage 2 calculation units M21, which are provided to the calculation units M22. Figure 8 shows that 16 input signals are input to 8 calculation units M21, the minimum values calculated by every 2 calculation units M21 are input to the next stage 4 calculation units M21, and the first minimum value (minl) and the pseudo second minimum value (min2'") are obtained by the next stage calculation units M42.

[0068] Taking the check node degree of 16 as an example, the first stage 16 input signals are variable nodes (numbered n) to provide information (v2c information) to check nodes (numbered m), which can be expressed as ​The first minimum value (minl) and the associated information can then be obtained by the circuit shown in the figure, and the pseudo second minimum value (min2"') can be said to be the real second minimum value plus noise (scrambled second minimum value).

[0069] The decoding method and decoding system based on weight adjustment algorithm parameters disclosed in the present disclosure are improvements to the conventional single minimum and minimum sum algorithm (SMAMSA). The estimated second minimum value is modified to Equation 6, and Equation 6 is referred to as the modified single minimum and minimum sum algorithm (M-SMAMSA).

[0070] In Equation 6, the function is a minimum function, and in this example is the information from the variable node to the check node The minimum value (excluding the link to be calculated) can be used to obtain the estimated first minimum value and the estimated second minimum value, according to the information from the variable node to the check node (excluding the link to be calculated, and calculating the inner product (П) of the remaining links (n')), that is, estimating the information of itself according to the adjacent links, and then returning the information from the check node to the variable node .

[0071] Equation 6:

[0072] For m∈{1,…M}and n∈N m ;

[0073]

[0074]

[0075]

[0076] min1 est = a · minl;

[0077] min2 est = β · minl + γ · min2" '.

[0078] In comparison with the conventional single minimum and minimum sum algorithm, the estimated first minimum value (minl est ) obtained by the modified single minimum and minimum sum algorithm in Equation 6 is still consistent with the first minimum value obtained by the unmodified Equation 5, where a (the first parameter) is a weight value, and the estimated second minimum value (min2 estWhen multiplying by the first minimum value (min1) α and the pseudo second minimum value (min2″′) γ (third parameter), the operation parameter β (second parameter) multiplied by the first minimum value (min1) is also added.

[0079] Thus, the modified single minimum minimum sum algorithm changes the estimation of the second minimum (min2). est The modified single minimum sum algorithm, which originally had two dimensions (e.g., Equation 5) and three dimensions (e.g., Equation 6), expands its applicability to include more modulation schemes and a wider range of fixed points. According to the embodiment, when using the modified single minimum sum algorithm, it is important to note that Equation 6, after incorporating the variables α, β, and γ, must conform to the following rules.

[0080] min2 est ≥min1 est ;

[0081] Since min2″′ ≥ min1, min2″′ is at least equal to min1, and min2″′ is at least equal to min1. est The lowest bound is when min2″′=min1, from which min2 can be derived. est = (β+γ)·min1≥min1 est =α·min1, therefore, when giving the parameters α, β and γ, the relation (β+γ)≥α must be satisfied.

[0082] For an example of a logic circuit block diagram for implementing the modified single minimum sum algorithm, please refer to [reference needed]. Figure 9 The decoding method based on the weight adjustment algorithm parameters can be found in the following reference. Figure 10 The illustrated implementation flow.

[0083] exist Figure 10 In the displayed process, after receiving the signal, the decoder generates an M×N low-density parity check code from the input signal, which has N variable nodes and M check nodes, resulting in multiple input signals 901 and 902 (step S101). The input signals are information from multiple variable nodes to multiple check nodes. When calculating the update of the verification node, Equation 6 is applied, using the calculation unit 90 shown in the diagram (see embodiment for reference). Figures 5 to 8 Calculate the first minimum value (min1) (step S103), and use the supplementary information obtained when calculating the first minimum value to obtain the pseudo second minimum value (min2″′) (step S105).

[0084] According to Equation 6 above, under the requirement of (β+γ)≥α, the estimated first minimum value is obtained by multiplying the first parameter (α) by the first minimum value. est (Step S107) The second parameter (β) is multiplied by the first minimum value and then multiplied by the third parameter (γ) by the pseudo-second minimum value to obtain the estimated second minimum value (min2). est (Step S109). At this point, based on the information provided by the variable node to the verification node, after excluding the connections to be calculated, the remaining connections are summed and then judged to be 0 or 1. (Step S111), then based on the judgment Is n (the number of the variable node) a label? The variable node n needs to provide information to the node with the number m as a verification node. The connection containing the minimum value (the variable node numbered n) determines whether to use the estimated first minimum value (min1). est Alternatively, if the above conditions are not met, the estimated second minimum value (min2) is used. est The inner product of the value (0 or 1) obtained from step S111 is calculated to obtain the information from the check node to the variable node. (Step S113).

[0085] Before a signal enters the decoder described in the above embodiments, preprocessing such as decode scaling can be performed. The purpose is to adjust the signal to a state where the decoder can recognize its features under hardware constraints. One method is to determine the scaling by comparing the noise power of the signal with a threshold set by the system, thereby improving the performance reduction problem caused by the bandwidth limitation of fixed points.

[0086] For example, the log-similarity ratio (LLR) entering a low-density parity-check (LDPC) decoder can be controlled by decoding scaling because the decoder requires the inverse of noise power (1 / σ) when performing the log-similarity ratio calculation. 2 The information is used to adjust the weighting of the input log-similarity ratio. The reciprocal of the noise power provides the signal-to-noise ratio (SNR) information. The strength of the signal in each channel is adjusted by the reciprocal of the noise power to provide the decoder with the correct log-similarity ratio.

[0087] In practical applications, because the signal-to-noise ratio range of the actual decoding system operation can span a range of 8-9 dB, the dynamic range of the noise power required by the low-density parity-check code decoder when inputting the log-likelihood ratio (LLR) falls within the range of 2 to 8. In order to completely cover the entire dynamic range, the fixed point of the decoder needs to increase the bit width to maintain the performance of the decoder, but this will increase the hardware area and power consumption. Therefore, a decoding scaling method for the log-likelihood ratio is proposed to suppress the change of the fixed point, including reducing the bit width of the fixed point.

[0088] In summary, according to the above decoding method and decoding system based on the weight adjustment algorithm parameter, the conventional single minimum value minimum sum algorithm is modified, and a new single minimum value architecture is combined with layered decoding technology to provide more dimensional variables than conventional algorithms, increase the application range, and provide better performance than the conventional normalized minimum sum algorithm (NMS). In addition, because the single minimum value is another scrambled second minimum value, a lower error rate can be provided. Furthermore, because only the first minimum value needs to be searched, the hardware complexity and power consumption can be reduced. At the input signal end, different decoding scaling methods are combined according to different signal-to-noise ratios, further reducing the hardware area and bit width, and providing more optimized performance.

[0089] The above disclosure is only a preferred embodiment of the present application, and does not limit the scope of the patent application. Any equivalent technical changes made in accordance with the content of the present application specification and drawings are included in the scope of the patent application.

[0090] Symbol explanation

[0091] 101: input circuit

[0092] 103: LLR operator

[0093] 105: low-density parity-checker

[0094] 107: output circuit

[0095] 21: variable node

[0096] 22: check node

[0097] 211: nth variable node

[0098] 221: mth check node

[0099] 201: v2c information

[0100] 202: c2v information

[0101] 311, 312, 313: check node

[0102] 301, 302, 303: wire

[0103] 411, 412, 413: variable node

[0104] 401, 402, 403: wire

[0105] M41, M42, M21, M22: calculation unit

[0106] 901, 902: input signal

[0107] 90: calculation unit

[0108] min1: first minimum value

[0109] min2": pseudo second minimum value

[0110] a, b, g: parameter

[0111] min1 est : estimated first minimum value

[0112] min2 est : estimated second minimum value

[0113] steps S101-S113: decoding method flow based on weight adjustment algorithm parameter

Claims

1. A decoding method based on weight adjustment algorithm parameters, applied in a decoder, wherein the input signal generates an M×N low-density parity check code, which includes multiple variable nodes and multiple check nodes, wherein the multiple variable nodes include N variable nodes and the multiple check nodes include M check nodes, and the decoding method based on weight adjustment algorithm parameters includes: Initialize the information of the multiple variable nodes and the multiple verification nodes; Update the multiple variable nodes, where each variable node is updated based on information from multiple interconnected check nodes. This process involves multiple iterations to generate information from each variable node to the multiple check nodes. After excluding the connections to be calculated, the remaining connections are summed. The multiple verification nodes are updated, with each node updating based on information from multiple interconnected variable nodes. Through multiple iterations, information from each verification node to the multiple variable nodes is generated. After excluding connections to be calculated, the remaining connections are multiplied. Then, based on the judgment and estimation of the first or second minimum value, the inner product is calculated to obtain the information from the verification node to the variable node, and a decision is executed accordingly. Search for the minimum value of the updated multiple variable nodes to obtain the first minimum value; The pseudo-second minimum value is derived by utilizing the additional information obtained when the first minimum value was determined. The estimated first minimum value is obtained by multiplying the first parameter (α) by the obtained first minimum value. The estimated second minimum is obtained by multiplying the first minimum by the second parameter (β) and adding the result of multiplying the pseudo second minimum by the third parameter (γ).

2. The decoding method based on weight adjustment algorithm parameters as described in claim 1, wherein in the initialization step, before iteration is performed, the internal information of the multiple variable nodes is written one by one.

3. The decoding method based on weight adjustment algorithm parameters as described in claim 1, wherein the method for determining the estimated first minimum value or the estimated second minimum value is determined based on the judgment result of whether the variable node number in the information given by the verification node to the variable node is the position of the information given by the smallest variable node to the verification node.

4. The decoding method based on weight adjustment algorithm parameters as described in claim 1, wherein, The decision is made by summing the intrinsic information of the variable node and the information updated to multiple variable nodes through the connection between the check node and other variable nodes.

5. The decoding method based on weight adjustment algorithm parameters as described in claim 1, wherein in the preprocessing of the decoder, the logarithmic similarity ratio is controlled by decoding scaling to adjust the weight value of the input logarithmic similarity ratio.

6. The decoding method based on weight adjustment algorithm parameters as described in any one of claims 1 to 5, wherein the first parameter (α), the second parameter (β), and the third parameter (γ) satisfy the relation (β+γ)≥ α.

7. The decoding method based on weight adjustment algorithm parameters as described in claim 6, wherein the information equation from the verification node to the variable node is: For ; ； min1; ； in, 'n' is the variable node number; 'm' is the check node number; 'Provide the information for the check node numbered m to the variable node numbered n;' 'Indicates the node numbers of the remaining variables in the connection to be calculated, excluding them;' 'Excluding the remaining numbers after excluding the links to be calculated' The variable node provides information to the verification node with number m, which is the V2C information; the symbolic function 'sign()' returns 0, 1, or -1 depending on whether the value in the function is 0, positive, or negative. 'The first minimum value; function' 'A function that takes the minimum value;' 'To estimate the first minimum value;' 'To estimate the second minimum value;' 'This is the pseudo-second minimum value.

8. A decoding system, the decoding system comprising a decoder located at a receiving end, wherein a decoding method based on weight adjustment algorithm parameters is executed, comprising: The input signal generates an M×N low-density parity check code, which has multiple variable nodes and multiple check nodes. The multiple variable nodes include N variable nodes and the multiple check nodes include M check nodes. Initialize the information of the multiple variable nodes and the multiple verification nodes; Update the multiple variable nodes, where each variable node is updated based on information from multiple interconnected check nodes. This process involves multiple iterations to generate information from each variable node to the multiple check nodes. After excluding the connections to be calculated, the remaining connections are summed. The multiple verification nodes are updated, with each node updating based on information from multiple interconnected variable nodes. Through multiple iterations, information from each verification node to the multiple variable nodes is generated. After excluding connections to be calculated, the remaining connections are multiplied. Then, based on the judgment and estimation of the first or second minimum value, the inner product is calculated to obtain the information from the verification node to the variable node, and a decision is executed accordingly. Search for the minimum value of the updated multiple variable nodes to obtain the first minimum value; The pseudo-second minimum value is derived by utilizing the additional information obtained when the first minimum value was determined. The estimated first minimum value is obtained by multiplying the first parameter (α) by the obtained first minimum value. The estimated second minimum is obtained by multiplying the first minimum by the second parameter (β) and adding the result of multiplying the pseudo second minimum by the third parameter (γ).

9. The decoding system of claim 8, wherein the method for determining the estimated first minimum value or the estimated second minimum value is determined based on the determination result of whether the variable node number in the information given by the verification node to the variable node is the position of the information given by the smallest variable node to the verification node.

10. The decoding system as described in claim 8, wherein, The decision is made by summing the intrinsic information of the variable node and the information updated to multiple variable nodes through the connection between the check node and other variable nodes.

Images