A shortened LDPC decoding method and system with high coding gain

Abstract

The application discloses a shortened LDPC decoding method and system with high coding gain, and the method comprises the following steps: initializing channel reception log-likelihood ratio information and extrinsic information transmitted from check nodes to variable nodes; taking the initialized channel reception log-likelihood ratio information and the extrinsic information transmitted from the check nodes to the variable nodes as first iteration input, performing iteration decoding calculation to obtain a code word decision result; and outputting the code word decision result as a decoding result. The application improves the accuracy of decoding and the convergence speed of decoding.

Description

Technical Field

[0001] This invention belongs to the technical field of LDPC code decoding methods, and particularly relates to a shortened LDPC decoding method and system with high coding gain. Background Technology

[0002] The (8176, 7154) LDPC code of the CCSDS near-Earth standard is widely used in my country's satellite communication field due to its excellent characteristics such as low coding overhead, fast iterative convergence speed, and relatively simple encoding and decoding implementation. However, its 8176-bit code length poses a significant limitation for common applications such as parallel high-speed processing and transmitting integer frame data per unit time. Therefore, the practical application of this code is usually accompanied by shortening. For example, the CCSDS frame format uses a shortened code with 18 bits of virtual padding, resulting in a code length of 8160 bits; my country's laser satellite communication uses a shortened code with 210 bits of virtual padding, resulting in a code length of 7966 bits. Including the two padding zero bits, the total data block length is 7968 bits. Although the shortening operation can effectively match the codeword length, since the virtual padding bits do not participate in the actual data transmission and reception, the effective time of the decoding iteration process is shortened. The reduction in the number of decoding iterations will significantly degrade the decoding performance. When the number of shortened bits is small, such as in applications using the CCSDS frame format, the problem of shortened decoding time can be eliminated by setting the bits at the end of the frame synchronization header that are the same as the shortened length to zero, thus supplementing the shortened codeword to the original length. However, for the 210-bit shortened code used in laser satellite communication, its shortened length is much greater than the 32 bits of the frame synchronization header, so the problem of shortened decoding time cannot be eliminated by supplementing the code length with synchronization header bits. Summary of the Invention

[0003] The technical problem solved by this invention is to overcome the shortcomings of the prior art and provide a shortened LDPC decoding method and system with high coding gain, which improves the decoding accuracy and decoding convergence speed.

[0004] The objective of this invention is achieved through the following technical solution: a shortened LDPC decoding method with high coding gain, comprising: initializing the channel received log-likelihood ratio information and the extrinsic information transmitted by the check node to the variable node; using the initialized channel received log-likelihood ratio information and the extrinsic information transmitted by the check node to the variable node as the first iteration input, performing iterative decoding calculation to obtain the codeword decision result; and outputting the codeword decision result as the decoding result.

[0005] In the above-mentioned high coding gain shortened LDPC decoding method, the initial channel receive log-likelihood information includes: the initial channel receive log-likelihood information for virtual padding bits and the initial channel receive log-likelihood information for other transmitted information bits.

[0006] In the aforementioned high coding gain shortened LDPC decoding method, the channel received log-likelihood ratio information f n When the variable node number n is 0≤n≤S-1, the channel receives the log-likelihood ratio information f. n The channel receives log-likelihood ratio information for virtual padding bits; where S is the shortened code length; the channel receives log-likelihood ratio information f. n When the variable node number n is S≤n≤N-1, the channel receives the log-likelihood ratio information f. n The channel receives the log-likelihood ratio information for other transmitted information bits; where S is the shortened code length and N is the code length.

[0007] In the aforementioned high coding gain shortened LDPC decoding method, the channel received log-likelihood ratio information f n The initial value is determined by the virtual padding identifier v. n The ratio of the log-likelihood of the received channel to the channel g n It consists of two parts, v n Taking 1 indicates that f n The corresponding bit is the virtual padding bit, v n Taking 0 indicates that f n The corresponding bit is the normal transmission bit, g n Indicates that f n The corresponding log-likelihood ratio of 0 and 1 for the received bits.

[0008] In the above-mentioned shortened LDPC decoding method with high coding gain, the extrinsic information L passed from the initialization check node to the variable node is... m→n For the external information L passed from the verification node to the variable node m→n Set to 0; where 0≤m≤W-1, 0≤n≤N-1, W is the length of the code's check bits, N is the code length, m is the check node number, and n is the variable node number.

[0009] In the aforementioned shortened LDPC decoding method with high coding gain, the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node are used as the initial iteration inputs for iterative decoding calculations. The resulting codeword decision includes: obtaining the codeword bit decision value based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node; determining whether to continue iterative calculation based on the current iteration number and the codeword bit decision value; and if iterative calculation is required, obtaining the extrinsic information value Z transmitted from the variable node to the check node based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. n→m Based on the extrinsic information passed from the variable node to the check node, the new extrinsic information passed from the check node to the variable node is obtained.

[0010] In the aforementioned high-coding-gain shortened LDPC decoding method, the codeword bit decision value Z is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. n Includes: when the channel receives log-likelihood ratio information f n The virtual fill identifier v in n When it is 1:

[0011] Z n =0;

[0012] When the channel receives log-likelihood ratio information f n The virtual fill identifier v in n When it is 0:

[0013]

[0014] Where sign is the sign decision operation, L m→n To verify the external information passed from the node to the variable node, This is the set of check nodes connected to the variable node index n.

[0015] In the aforementioned high-gain shortened LDPC decoding method, determining whether to continue iterative calculation includes: checking whether the iteration number it is equal to the preset maximum iteration number IT. max If it = IT max If the iteration stops, then the decision codeword C is set to 1. code Substitute the LDPC check matrix H into the check matrix for verification calculation. If C code ·H T =0 indicates that the decoding result satisfies the check equation, and the iteration stops; if cH T If it is not 0, increment the iteration count it by 1 and perform iterative calculation.

[0016] In the above-mentioned shortened LDPC decoding method with high coding gain, the extrinsic information value Z transmitted from the variable node to the check node is obtained. n→m Includes: when the channel receives log-likelihood ratio information f n The virtual fill identifier v in n When it is 1:

[0017] Z n→m =G max ;

[0018] When the channel receives log-likelihood ratio information f n The virtual fill identifier v in n When it is 0:

[0019]

[0020] Among them, g n Indicates that fn The corresponding log-likelihood ratio of 0 and 1 of the received bits. This is the set of check nodes connected to the variable node index n. Indicates from set G is the subset after removing element m. max For g n The largest positive integer that can be represented within the quantization range.

[0021] In the aforementioned shortened LDPC decoding method with high coding gain, the new check node transmits extrinsic information L' to the variable node. m→n It can be obtained through the following formula:

[0022]

[0023] Among them, Z n→m This refers to the external information value passed from the variable node to the verification node. Indicates from set The subset after removing element n. This represents the set of check nodes connected to the variable node index n.

[0024] A high coding gain shortened LDPC decoding system includes: a first module for initializing channel reception log-likelihood ratio information and extrinsic information transmitted from the check node to the variable node; a second module for performing iterative decoding calculations using the initialized channel reception log-likelihood ratio information and extrinsic information transmitted from the check node to the variable node as the first iteration input to obtain codeword decision results; and a third module for outputting the codeword decision results as the decoding result.

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] (1) In the LDPC decoding calculation, the present invention locks the decision value of the virtual stuffing bit to 0, so as to prevent the value of the virtual stuffing bit from being flipped due to the error information of other nodes during the iterative correction of the decoding information, and further transmits it to other nodes, thereby improving the accuracy of decoding.

[0027] (2) By always setting the external information transmitted by the virtual filling bit to the maximum positive value, the present invention maximizes the correction capability of the prior information in the iterative decoding process and improves the convergence speed of decoding. Attached Figure Description

[0028] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0029] Figure 1 This is a schematic diagram of the composition structure of the (7966, 6944) LDPC codeword provided in an embodiment of the present invention;

[0030] Figure 2 This is a schematic diagram illustrating the iterative decoding performance provided in an embodiment of the present invention. Detailed Implementation

[0031] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0032] Laser satellite communication uses (7966, 6944) LDPC encoding, which is formed by virtual padding the first 210 bits of (8176, 7154) LDPC encoding. Its composition is shown in the attached figure. Figure 1 As shown. The transmitting end does not send the 210-bit virtual padding bits before encoding, and the receiving end removes the last 2 bits of padding bits before decoding. The LDPC check area is 128 bytes.

[0033] This embodiment provides a shortened LDPC decoding method with high coding gain, which includes the following steps:

[0034] Step 1: Initialize the channel reception log-likelihood ratio information and the external information transmitted from the check node to the variable node. Initializing the channel reception log-likelihood ratio information includes: initializing the channel reception log-likelihood ratio information for virtual padding bits and initializing the channel reception log-likelihood ratio information for other transmitted information bits.

[0035] Specifically, it is divided into three initialization processes: A), B), and C).

[0036] A) Initialize the channel receive log-likelihood ratio information for the virtual padding bits. The log-likelihood ratio information of the nth bit is represented by f. n This indicates that 0 ≤ n ≤ S-1, where S is the shortened length of the code. n The initial value is determined by the virtual padding identifier v. n The ratio of the log-likelihood of the received channel to the channel g n It consists of two parts, v n Taking 1 indicates that f n The corresponding bit is the virtual padding bit, v nTaking 0 indicates that f n The corresponding bit is the normal transmission bit, g n Indicates that f n The corresponding log-likelihood ratio of the received bits (0 and 1) has the same meaning as the input likelihood ratio in a conventional log-likelihood decoding algorithm. The f-value of the virtual padding bits... n The initialization process involves the following two steps: First, v n Set bit 1; second, set g n Let's set it to G max G max For g n The largest positive integer that can be represented within the quantization range.

[0037] B) Initialize the channel receive log-likelihood ratio information f for other transmitted information bits. n In this case, S≤n≤N-1, where S is the shortened length of the code and N is the code length. The f values ​​of other transmitted information bits... n The initialization process involves the following two steps: First, v n Set the bit to 0; second, calculate g. n And assign values, the specific calculation relationship is g n =log(P(x) n =0|y n ) / P(x n =1|y n )), where P(x n =0|y n ) represents the received level value of the nth bit. n Send bit x at time n The probability that x = 0, P(x) n =1|y n ) represents the received level value of the nth bit. n Send bit x at time n The probability of the received level being 1. Based on the likelihood function of the Gaussian channel, the probability of the received level value of the nth bit is calculated as follows:

[0038]

[0039]

[0040] Where, σ 2 This represents the variance of the channel's Gaussian white noise.

[0041] C) Initialize the external information values ​​passed from all check nodes to the variable nodes, L m→n L represents the extrinsic information value passed from the check node with index m to the variable node with index n, where 0 ≤ m ≤ W-1, 0 ≤ n ≤ N-1, and W is the length of the check bits in the code. m→n The initialization process is to initialize all Lm→n Set to 0.

[0042] Step 2: Set the iteration count it to 0 to initialize the obtained f. n and L m→n The data value is the input for the first iteration. Iterative decoding calculation is performed to obtain the codeword decision result, which is specifically divided into four steps: a), b), c), and d).

[0043] a) Calculation of codeword decision value:

[0044] For each bit of information in the codeword, it is represented by f corresponding to the nth bit. n and L m→n Z was calculated n Z n It is the decision value of the nth bit in the codeword, where 0≤n≤N-1, and the specific process is as follows:

[0045] When the channel received by variable node n has a log-likelihood ratio of f n The virtual fill identifier v in n When it is 1:

[0046] Z n =0

[0047] When the channel received by variable node n has a log-likelihood ratio of f n The virtual fill identifier v in n When it is 0:

[0048]

[0049] in, This represents the set of check nodes connected to variable node n. `sign` is the sign decision operation. When the input value is greater than or equal to zero, the output is 0; otherwise, the output is 1.

[0050] b) Iteration termination determination:

[0051] Based on the current iteration number it and the codeword bit decision value Z n The process of determining whether to continue iterative calculation is as follows:

[0052] Check if the current iteration number it equals the preset maximum iteration number IT. max If it = IT max If the iteration stops, proceed to step three to output the decoding result; otherwise, output the decision codeword C. code Substitute the LDPC check matrix H into the check matrix for check calculation, where the decision codeword C code The decision value Z of each bit n Composition, including C code =[Z0,Z1,…,Z j,…,Z N-1 If C code • HT = 0 indicates that the decoding result satisfies the check equation, the iteration stops, and the process proceeds to step three to output the decoding result. If cH T If it is not 0, increment the iteration count it by 1 and proceed to c) for a new iteration.

[0053] c) Calculation of variable node updates:

[0054] For each variable node n and From the initial f n and L m→n The extrinsic information value Z transmitted from variable node n to check node m is calculated. n→m ,in This represents the set of check nodes connected to the variable node n. The specific process is as follows:

[0055] When the channel received by variable node n has a log-likelihood ratio of f n Virtual fill identifier v n When it is 1:

[0056] Z n→m =G max

[0057] When the channel received by variable node n has a log-likelihood ratio of f n Virtual fill identifier v n When it is 0:

[0058]

[0059] Among them, g n It is f n The corresponding log-likelihood ratio of 0 and 1 of the received bits. Indicates from set The subset after removing element m.

[0060] d) The verification node updates and calculates the new external information passed from the verification node to the variable node:

[0061] For each verification node m and Z obtained from c) n→m Calculate the extrinsic information L' passed from the new check node to the variable node in this iteration. m→n , This represents the set of check nodes connected to the variable node n. Indicates from set The subset after removing element n, and the external information L passed from the verification node to the variable node. m→n The update calculation formula is as follows:

[0062]

[0063] Z n→m This refers to the external information value passed from variable node n to verification node m.

[0064] L' m→n After the value is updated, return to step a) in step two to perform code word judgment for this iteration.

[0065] Step 3: Output the decoding result:

[0066] The judgment code C code Output as the decoding result.

[0067] The following example, using a (7966, 6944) LDPC code in laser satellite communication, illustrates the specific implementation of this invention. In this example, the channel received likelihood ratio (LLR) information is quantized using 6 bits, and the range of the quantity and value is [-32, 31], with a maximum of 6 iterations. The specific implementation steps are as follows:

[0068] Step 1: Initialize the channel to receive log-likelihood ratio information and extrinsic information transmitted from the check node to the variable node. This is specifically divided into three initialization processes: A, B, and C.

[0069] A) Initialize the channel receive log-likelihood ratio information for the virtual padding bits. The log-likelihood ratio information of the nth bit is represented by f. n This indicates that 0 ≤ n ≤ 209. n The initial value is determined by the virtual padding identifier v. n The ratio of the log-likelihood of the received channel to the channel g n It consists of two parts, v n Taking 1 indicates that f n The corresponding bit is the virtual padding bit, v n Taking 0 indicates that f n The corresponding bit is the normal transmission bit, g n Indicates that f n The corresponding log-likelihood ratio of the received bits (0 and 1) has the same meaning as the input likelihood ratio in a conventional log-likelihood decoding algorithm. The f-value of the virtual padding bits... n The initialization process involves the following two steps: First, v n Set bit 1; second, set g n Set it to 31.

[0070] B) Initialize the channel receive log-likelihood ratio information f for other transmitted information bits. n At this point, 2^10 ≤ n ≤ 7965. The f values ​​for other transmitted information bits... n The initialization process involves the following two steps: First, v n Set to 0;

[0071] Second, calculate g n And assign values, the specific calculation relationship is as follows:

[0072] g n =log(P(x) n =0|y n ) / P(x n =1|y n ));

[0073] Wherein, P(x n =0|y n ) represents the received level value of the nth bit. n Send bit x at time n The probability that x = 0, P(x) n =1|y n ) represents the received level value of the nth bit. n Send bit x at time n The probability of the received level being 1. Based on the likelihood function of the Gaussian channel, the probability of the received level value of the nth bit is calculated as follows:

[0074]

[0075]

[0076] Where, σ 2 This represents the variance of the channel's Gaussian white noise.

[0077] C) Initialize the external information values ​​passed from all check nodes to the variable nodes, L m→n This represents the extrinsic information value passed from the check node with index m to the variable node with index n, where 0 ≤ m ≤ 1021 and 0 ≤ n ≤ 7965. m→n The initialization process is to initialize all L m→n Set to 0.

[0078] Step 2: Set the iteration count it to 0 to initialize the obtained f. n and L m→n The data value is input for the first iteration. Iterative decoding calculation is performed to obtain the codeword decision result, which is divided into four steps: a, b, c, and d.

[0079] a) Calculation of codeword decision value:

[0080] For each bit of information in the codeword, it is represented by f corresponding to the nth bit. n and L m→n Z was calculated n Z n It is the decision value of the nth bit in the codeword, where 0≤n≤7965. The specific process is as follows:

[0081] When the channel received by variable node n has a log-likelihood ratio of f n The virtual fill identifier v in n When it is 1:

[0082] Z n =0

[0083] When the channel received by variable node n has a log-likelihood ratio of f n The virtual fill identifier v in n When it is 0:

[0084]

[0085] The sign operator is used for sign decision. When the input value is greater than or equal to zero, the output is 0; otherwise, the output is 1.

[0086] b) Iteration termination determination:

[0087] Based on the current iteration number it and the codeword bit decision value Z n The process of determining whether to continue iterative calculation is as follows:

[0088] Check if the current iteration number *it* equals 6. If *it* = 6, stop the iteration and proceed to step three to output the decoding result; otherwise, output the decision codeword C. code Substitute the LDPC check matrix H into the check matrix for check calculation, where the decision codeword C code The decision value Z of each bit n Composition, including C code =[Z0,Z1,…,Z j ,…,Z 7965 ].like This indicates that the decoding result satisfies the check equation, the iteration stops, and the process proceeds to step three to output the decoding result. If cH T If it is not 0, increment the iteration count it by 1 and proceed to c) for a new iteration.

[0089] c) Calculation of variable node updates:

[0090] For each variable node n and From the initial f n And L in the previous iteration m→n The extrinsic information value Z transmitted from variable node n to check node m is calculated. n→m ,in This represents the set of check nodes connected to the variable node n. The specific process is as follows:

[0091] When the channel received by variable node n has a log-likelihood ratio of f n Virtual fill identifier v n When it is 1:

[0092] Z n→m =G max

[0093] When the channel received by variable node n has a log-likelihood ratio of f n Virtual fill identifier v n When it is 0:

[0094]

[0095] Where g n It is f n The corresponding log-likelihood ratio of 0 and 1 of the received bits. Indicates from set The subset after removing element m.

[0096] d) Verification node update calculation:

[0097] For each verification node m and Z obtained from c) n→m Calculate L' in this iteration n→m value, Let n be the set of check nodes connected to variable node n. The calculation formula is as follows:

[0098]

[0099] in, Indicates from set The subset after removing element n.

[0100] L' n→m After the value is updated, return to step a) in step two to perform code word judgment for this iteration.

[0101] Step 3: Output the decoding result:

[0102] The judgment code C code Output as the decoding result.

[0103] This invention simulates the performance of the (7966, 6944) LDPC code used in laser satellite communication using the proposed decoding method. The simulation was performed with 6 iterations, and the results are shown in the appendix. Figure 2 As shown.

[0104] From the appendix Figure 2As can be seen, the method proposed in this invention is used to decode the shortened (7966, 6944) LDPC code in 6 iterations. At the same time, the minimum sum decoding method is used to decode the unshortened (8176, 7154) LDPC code in 7 iterations. Although only 6 decoding iterations can be performed due to the shorter processing time of the shortened codeword, the method proposed in this invention makes full use of the prior information of the virtual padding bits, and the iteration convergence is faster. The performance is comparable to the 7-iteration decoding of the unshortened code. The 6-iteration decoding of the unshortened code has a significantly reduced performance of about 1dB due to the reduction in the number of iterations.

[0105] This embodiment also provides a high coding gain shortened LDPC decoding system, which includes: a first module for initializing channel reception log-likelihood ratio information and extrinsic information passed from the check node to the variable node; a second module for performing iterative decoding calculations using the initialized channel reception log-likelihood ratio information and extrinsic information passed from the check node to the variable node as the first iteration input to obtain codeword decision results; and a third module for outputting the codeword decision results as decoding results.

[0106] This invention addresses the problem of degraded decoding performance caused by the large number of shortened bits in the (7966, 6944) LDPC code used in laser satellite communication, which leads to shortened decoding iteration processing time. By effectively utilizing the prior information of virtual padding bits, known node information is transmitted to other nodes with high reliability, accelerating the iteration convergence speed of other nodes. Therefore, excellent decoding performance can be maintained even with a shortened iteration processing time. This invention is not limited to the (7966, 6944) LDPC code used in laser satellite communication, but can also be applied to other shortened LDPC codes with longer shortened lengths.

[0107] This embodiment improves decoding accuracy by locking the decision value of the virtual stuffing bit to 0 during LDPC decoding calculation. This prevents the value of the virtual stuffing bit from being flipped due to erroneous external information from other nodes during iterative correction of decoding information, and further transmitting it to other nodes. This embodiment also maximizes the corrective ability of prior information in the iterative decoding process by always setting the external information transmitted by the virtual stuffing bit to the maximum positive value, thereby improving the convergence speed of decoding.

[0108] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A shortened LDPC decoding method with high coding gain, characterized in that... include: Initialize the channel to receive log-likelihood ratio information and extrinsic information passed from the check node to the variable node; Using the channel received log-likelihood ratio information obtained during initialization and the extrinsic information passed from the check node to the variable node as the first iteration input, iterative decoding calculation is performed to obtain the code word decision result; The code word judgment result is output as the decoding result; Using the channel received log-likelihood ratio information obtained during initialization and the extrinsic information passed from the check node to the variable node as the input for the first iteration, iterative decoding calculation is performed to obtain the codeword decision results, including: The codeword bit decision value is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. Based on the current iteration count and codeword bit decision value, determine whether to continue iterative calculation; If iterative calculation is required, the extrinsic information value transmitted from the variable node to the check node is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. Based on the extrinsic information passed from the variable node to the check node, the new extrinsic information passed from the check node to the variable node is obtained. The codeword bit decision value is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. include: When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 1: ； When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 0: ； in, sign For symbolic decision operations, To verify the external information passed from the node to the variable node, For the variable node number n A set of connected verification nodes; Determining whether to continue iterative calculations includes: Check the number of iterations it Is it equal to the preset maximum number of iterations? IT max ,like it=IT max If the iteration stops, then the decision codeword will be entered. Substitute into the LDPC check matrix Perform verification calculations, if This indicates that the decoding result satisfies the check equation, and the iteration stops; if The number of iterations it Increase by 1 and perform iterative calculations; Obtain the extrinsic information value passed from the variable node to the verification node. include: When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 1: ； When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 0: in, g n This indicates that f n The corresponding log-likelihood ratio of 0 and 1 of the received bits. For the variable node number n A set of connected verification nodes. Indicates from set Remove elements m Subsequent subsets, G max for g n The largest positive integer that can be represented within the quantization range; External information passed from the new verification node to the variable node It can be obtained through the following formula: ； in, This refers to the external information value passed from the variable node to the verification node. Indicates from set Remove elements n Subsequent subsets, Indicates the node number of the variable. n A set of connected verification nodes.

2. The high coding gain shortened LDPC decoding method according to claim 1, characterized in that: The initial channel receive log-likelihood ratio information includes: the initial channel receive log-likelihood ratio information for virtual padding bits and the initial channel receive log-likelihood ratio information for other transmitted information bits.

3. The high coding gain shortened LDPC decoding method according to claim 2, characterized in that: Channel receive log-likelihood ratio information f n Variable node number in n 0≤ n ≤ S When -1, the channel receives the log-likelihood ratio information. f n The channel receives log-likelihood ratio information for virtual padding bits; in, S It is the shortened length of the code; Channel receive log-likelihood ratio information f n Variable node number in n for S ≤ n ≤ N When -1, the channel receives the log-likelihood ratio information. f n The channel receives the log-likelihood ratio information for other transmitted information bits; among which, S It is the shortened length of the code. N It is the code length; Channel receive log-likelihood ratio information f n The initial value is indicated by virtual padding. v n Ratio of the channel received log-likelihood g n It consists of two parts. v n Taking 1 indicates that f n The corresponding bits are virtual padding bits. v n Taking 0 indicates that f n The corresponding bits are normal transmission bits. g n This indicates that f n The corresponding log-likelihood ratio of 0 and 1 for the received bits.

4. The high coding gain shortened LDPC decoding method according to claim 1, characterized in that: The external information passed from the initialization check node to the variable node To verify the external information passed from the node to the variable node Set to 0; where 0 ≤ m ≤ W -1, 0≤ n ≤ N -1 ，W It is the length of the check bits in the code. N It's the code length. m To verify the node sequence number, n This is the node number of the variable.

5. A shortened LDPC decoding system with high coding gain, characterized in that... include: The first module is used to initialize the channel reception log-likelihood ratio information and the external information passed from the check node to the variable node; The second module is used to perform iterative decoding calculations using the channel reception log-likelihood ratio information obtained during initialization and the external information passed from the check node to the variable node as the first iteration input, and to obtain the code word decision result. The third module is used to output the code word decision result as the decoding result; Using the channel received log-likelihood ratio information obtained during initialization and the extrinsic information passed from the check node to the variable node as the input for the first iteration, iterative decoding calculation is performed to obtain the codeword decision results, including: The codeword bit decision value is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. Based on the current iteration count and codeword bit decision value, determine whether to continue iterative calculation; If iterative calculation is required, the extrinsic information value transmitted from the variable node to the check node is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. Based on the extrinsic information passed from the variable node to the check node, the new extrinsic information passed from the check node to the variable node is obtained. The codeword bit decision value is obtained based on the channel received log-likelihood ratio information and the extrinsic information transmitted from the check node to the variable node. include: When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 1: ； When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 0: ； in, sign For symbolic decision operations, To verify the external information passed from the node to the variable node, For the variable node number n A set of connected verification nodes; Determining whether to continue iterative calculations includes: Check the number of iterations it Is it equal to the preset maximum number of iterations? IT max ,like it=IT max If the iteration stops, then the decision codeword will be entered. Substitute into the LDPC check matrix Perform verification calculations, if This indicates that the decoding result satisfies the check equation, and the iteration stops; if The number of iterations it Increase by 1 and perform iterative calculations; Obtain the extrinsic information value passed from the variable node to the verification node. include: When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 1: ； When the channel receives log-likelihood ratio information f n Virtual fill identifier in v n When it is 0: in, g n This indicates that f n The corresponding log-likelihood ratio of 0 and 1 of the received bits. For the variable node number n A set of connected verification nodes. Indicates from set Remove elements m Subsequent subsets, G max for g n The largest positive integer that can be represented within the quantization range; External information passed from the new verification node to the variable node It can be obtained through the following formula: ； in, This refers to the external information value passed from the variable node to the verification node. Indicates from set Remove elements n Subsequent subsets, Indicates the node number of the variable. n A set of connected verification nodes.

Images