Adaptive LDPC encoding and decoding method with directional interaction, transmitting end, receiving end and system

Abstract

The present invention provides an adaptive LDPC encoding and decoding method with directional interaction, a transmitting end, a receiving end and a system. The method includes: the receiving end performs LDPC decoding and verification. If the verification fails, the following operations are executed: the LDPC decoding outputs confidence feature for confidence threshold decision, and the confidence features less than the threshold are cached; calculate the number of interaction information S and perform interaction information selection, feedback the information N of decoding verification failure and the interaction information identifier Pos, and the transmitting end selects the corresponding interaction content from the buffer according to the Pos position; after receiving the interaction information, the receiving end updates the soft information of the decoding end: update the soft information sequence LLR of the corresponding bit to the maximum value and decode again. Based on the decoding convergence situation of the LDPC code and the characteristics of the code matrix, the present invention triggers different interaction contents through qualitative and quantitative interaction information selection under different communication conditions, and realizes the adaptive adjustment of channel encoding and decoding.

Description

Technical Field

[0001] The present invention belongs to the technical field of channel coding in wireless communication, and particularly relates to a method for directional interaction-based adaptive LDPC encoding and decoding, a transmitting end, a receiving end, and a system. Background Art

[0002] LDPC (Low-density Parity-check) codes are linear block codes that approach the Shannon limit, have high coding gain, and can effectively reduce the transmission power of wireless devices. Under the same performance, LDPC decoders have a simpler structure, higher parallelism, and are more conducive to implementation compared to other code groups. The LDPC decoding process utilizes the information of channel error statistics, the prior probability of received information bits, iteratively outputs the posterior probability, and adopts a SISO (soft Input Soft Output) decoder. However, for the maximum likelihood decoding algorithm used, the decoder is prone to convergence failure under low signal-to-noise ratio conditions.

[0003] The solutions for optimizing performance based on the decoding output mainly focus on hierarchically quantifying the output bits of the decoder by using the magnitude (modulus value) of the soft decision log-likelihood ratio of the decoding output, determining the reliability of the decoding output result according to the determined level, and improving the performance by retransmitting the bits with low reliability selected according to the level, or variant solutions based on this, such as multiplying a quantization factor on the magnitude value of the soft decision log-likelihood ratio during the quantization level division, optimizing the level division, or directly dividing the error level by a decoding threshold in the error level part, and directly retransmitting the content under this threshold decision. The flowchart is as Figure 1 shown.

[0004] For improving the decoding performance by using the decoding soft information, it is necessary to specifically utilize the decision information after decoding iteration, fully consider the decoding convergence characteristics, such as the corresponding relationship between encoding and decoding, the decision basis of decoding probability information, and the complexity and overhead increased for improving decoding accuracy. Summary of the Invention

[0005] The present invention aims to solve the technical problems existing in the prior art, and provides a method for directional interaction-based adaptive LDPC encoding and decoding, a transmitting end, a receiving end, and a system, so as to solve the problems of low effectiveness of decoding feedback, limited improvement in transmission quality, and large resource overhead in the prior art.

[0006] According to the first aspect of the present invention, the present invention provides a method for directional interaction-based adaptive LDPC encoding and decoding, which includes the following steps:

[0007] S1. The sending end performs LDPC encoding on the information sequence to be encoded and caches it; the sending end transmits the LDPC - encoded data through a noisy channel to the receiving end.

[0008] S2. The receiving end receives the signal, performs LDPC decoding and LDPC decoding verification. If the decoded codeword can pass the decoding verification, it sends the correct feedback information Y to the sending end, indicating that the receiving end has completed the correct decoding of the received information. If it fails to pass the decoding verification, the following operations are performed:

[0009] a. Perform a confidence - threshold decision on the confidence - level characteristics of the soft - information sequence output by LDPC decoding, cache the confidence - level characteristics less than the threshold, and the quantity is D.

[0010] b. Calculate the number of mutual information S and perform mutual - information selection: If D >= S, then select the first S confidence - level characteristics in ascending order of confidence in the buffer, and use the corresponding bit positions Pos(S) as the orientation - information - sequence flag for feedback; if D < S, then feedback all the bit positions Pos(D) corresponding to the D confidence characteristics in the buffer.

[0011] c. Feedback the information N indicating the failure of decoding verification and the interactive information identifier Pos(S) / Pos(D). The sending end selects the corresponding interactive content from the buffer according to the Pos(S) / Pos(D) positions.

[0012] S3. After the receiving end receives the interactive information from the sending end, it updates the soft information at the decoding end: updates the LLR of the soft - information sequence of the corresponding bit to the maximum value and decodes again.

[0013] S4. Repeat the above steps until the decoding verification is satisfied or the maximum number of interactions is reached.

[0014] The adaptive LDPC encoding and decoding method with directional interaction of the present invention is based on the decoding convergence situation of LDPC codes and the characteristics of the code matrix. Under different communication conditions, through qualitative and quantitative selection of mutual information, different interactive contents are triggered to realize the adaptive adjustment of channel encoding and decoding. Based on the verification method of the LDPC parity - check matrix, the self - decision feedback of whether the decoding is successful is realized, reducing the time delay. Based on the confidence characteristics of the decoding information convergence of LDPC codes, an adaptive strategy for updating the decoding decision probability is adopted to improve the decoding accuracy. The qualitative and quantitative information caching and the selection strategy of dual - control of interactive information reduce the overhead and the burden of the feedback link.

[0015] According to a preferred embodiment of the present invention, the method of LDPC decoding verification is:

[0016] u=(u0,u1,…,u k-1) is the information sequence to be encoded, G is the generator matrix of the LDPC code, H is the parity-check matrix of the LDPC code, and the encoded codeword at the transmitter is c = u·G;

[0017] represents the corresponding codeword after iterative decoding at the receiver. After the LDPC decoding is completed, the codeword sequence and the parity-check matrix H satisfy the formula:

[0018] c·H T = 0 (1)

[0019] If then the check passes, otherwise the decoding fails.

[0020] The present invention realizes fast interactive feedback based on the self-determination of the code pattern matrix characteristics, reducing the time delay.

[0021] According to another preferred embodiment of the present invention, the method for calculating the confidence feature is:

[0022] Take the magnitude of the absolute value of the soft information sequence of each information bit output after iterative decoding of the LDPC code as the confidence feature, and the soft information sequence uses the log-likelihood ratio information.

[0023] According to still another preferred embodiment of the present invention, the confidence threshold threshold is a dynamic threshold (A) or a static threshold (B),

[0024] The dynamic threshold varies with the signal-to-noise ratio SNR of the transmission, and the static threshold reduces the threshold decision complexity;

[0025] where Rate is the code rate of the current LDPC code.

[0026] The present invention is based on the decoding confidence threshold, reduces the receiver buffer overhead and the calculation complexity, and accelerates the LDPC decoding convergence through qualitative (decoding confidence) and quantitative (dual control of the interaction quantity and the confidence threshold), improving the decoding accuracy.

[0027] According to still another preferred embodiment of the present invention, the calculation method of the single interaction information number S is:

[0028] It is determined by the algebraic relationship of the code rate Rate, the coding length Len, and the adaptive factor β of the current LDPC code:

[0029] S = Len×(0.028 - 0.02×Rate)×β (2)

[0030] Through the confidence threshold and the dynamic adjustment of the adaptive factor, the present invention calculates the single - interaction information number S, improves the transmission quality, reduces the cache overhead, and realizes the dynamic balance of broadband resources and error - correction capabilities, and the adaptive adjustment to meet different application requirements.

[0031] According to the second aspect of the present invention, a transmitting end is provided, which includes an encoding module, a cache module, and a directional information sending module; the encoding module receives the information to be encoded, performs LDPC encoding on the information - to - be - encoded sequence and caches it into the cache module, and the encoding module transmits the LDPC - encoded data through a noisy channel to the receiving end; the cache module receives the feedback decoding check - failure information N and the interaction information identifier Pos fed back by the receiving end, selects the corresponding interaction content from the cache according to the Pos position, and the interaction information identifier Pos is determined according to the adaptive LDPC encoding and decoding method of the directional interaction of the present invention.

[0032] The transmitting end of the present invention receives the feedback decoding check - failure information N and the interaction information identifier Pos fed back by the receiving end, selects the corresponding interaction content from the cache according to the Pos position, and triggers different interaction contents through qualitative and quantitative selection of interaction information under different communication conditions, so as to realize the adaptive adjustment of channel encoding and decoding.

[0033] According to the third aspect of the present invention, a receiving end is provided, which includes a decoding module, a checking module, a decision - making cache module, a directional interaction information calculation module, and a decoding information update module; the decoding module receives the encoded signal sent by the transmitting end, performs LDPC decoding and transmits it to the checking module and the decision - making cache module; the checking module performs decoding check. If the codeword after decoding can pass the decoding check, it sends the correct feedback information Y to the transmitting end, indicating that the receiving end has completed the correct decoding of the received information; if it fails to pass the decoding check, the directional interaction information calculation module performs the following operations: a. Perform a confidence - threshold decision on the confidence - level characteristics of the soft - information sequence output by LDPC decoding, cache the confidence - level characteristics less than the threshold, and the number is D; b. Calculate the interaction information number S and perform interaction information selection: if D>=S, then select the first S bit positions Pos(S) corresponding to the confidence - level characteristics in ascending order of confidence in the cache as the directional information sequence flag for feedback; if D<S, then feedback all the D confidence - characteristic - corresponding bit positions Pos(D) in the cache; c. Feed back the feedback decoding check - failure information N and the interaction information identifier Pos(S) / Pos(D) to the transmitting end; the decoding information update module receives the corresponding interaction content selected by the transmitting end according to the Pos(S) / Pos(D) position, updates the soft information at the decoding end, updates the soft - information sequence LLR of the corresponding bit to the maximum value and transmits it to the decoding module for re - decoding.

[0034] Based on the decoding convergence situation of LDPC codes and the characteristics of the code matrix, the receiving end of the present invention triggers different interaction contents through qualitative and quantitative mutual information selection under different communication conditions, realizing the adaptive adjustment of channel coding and decoding. Based on the checking method of the LDPC parity-check matrix, the present invention realizes the self-judgment feedback of whether the decoding is successful, reducing the time delay. Based on the confidence characteristics of the decoding information convergence of LDPC codes, the adaptive strategy for updating the decoding decision probability is improved, improving the decoding accuracy. The selection strategy of qualitative and quantitative information caching and dual control of mutual information reduces the overhead and the burden of the feedback link.

[0035] According to the fourth aspect of the present invention, the present invention provides a coding and decoding system, which includes the sending end of the present invention and the receiving end of the present invention.

[0036] Based on the decoding convergence situation of LDPC codes and the characteristics of the code matrix, the receiving end of the present invention triggers different interaction contents through qualitative and quantitative mutual information selection under different communication conditions, realizing the adaptive adjustment of channel coding and decoding. Based on the checking method of the LDPC parity-check matrix, the present invention realizes the self-judgment feedback of whether the decoding is successful, reducing the time delay. Based on the confidence characteristics of the decoding information convergence of LDPC codes, the adaptive strategy for updating the decoding decision probability is improved, improving the decoding accuracy. The selection strategy of qualitative and quantitative information caching and dual control of mutual information reduces the overhead and the burden of the feedback link.

[0037] The additional aspects and advantages of the present invention will be partly given in the following description, partly will become obvious from the following description, or will be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The above and / or additional aspects and advantages of the present invention will become obvious and easy to understand from the description of the embodiments in conjunction with the following drawings, wherein:

[0039] Figure 1 is a flowchart of a reliability-based retransmission strategy in the prior art;

[0040] Figure 2 is a flowchart of the adaptive LDPC coding and decoding method of directional interaction of the present invention;

[0041] Figure 3 is a flowchart of the adaptive LDPC coding and decoding method of directional interaction in a preferred embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0043] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "longitudinal", "lateral", "vertical", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the drawings. These are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention.

[0044] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it may be a mechanical connection or an electrical connection, or it may be the communication inside two elements. It may be directly connected or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0045] After in-depth analysis by the applicant, there are the following limitations for this type of solution that utilizes decoding soft information:

[0046] The statistical information of the LDPC decoding convergence situation and the code pattern characteristics are not fully utilized.

[0047] The incorrect data after the iterative decoding failure is not utilized. The retransmitted data lacks pertinence, the effectiveness of the transmitted information is low, and the ability to assist decoding is limited.

[0048] The feedback link has a large burden and the calculation complexity of the feedback information is high, and the cache cost is large.

[0049] Based on this, the present invention provides an adaptive LDPC encoding and decoding method with directional interaction, as Figure 2 and Figure 3 shown, which includes the following steps:

[0050] S1. The sending end performs LDPC encoding on the information sequence to be encoded and caches it; the sending end transmits the LDPC encoding through a noisy channel to the receiving end.

[0051] S2. The receiving end receives the signal, performs LDPC decoding, and then performs LDPC decoding verification. In this embodiment, the method of LDPC decoding verification is:

[0052] u=(u0,u1,…,u k-1 ,...) is the information sequence to be encoded, k is the information sequence number, G is the generating matrix of the LDPC code, H is the parity-check matrix of the LDPC code, and the encoded codeword at the sending end is c = u·G; Denote the corresponding codeword after iterative decoding at the receiving end. After the LDPC decoding is completed, the codeword sequence and the parity-check matrix H satisfy the formula:

[0053] c·H T =0 (1)

[0054] If then the check passes, otherwise the decoding fails.

[0055] Specifically, the generation matrix G and the parity-check matrix H can adopt the one-to-one corresponding matrix structure in the prior art, and the parity-check matrix and the generation matrix satisfy GH T =0.

[0056] For example: the information to be encoded: u = (u1 u2 u3), where u1, u2, and u3 are the information units to be encoded. After LDPC encoding, the codeword C(u) = (c1 c2 c3 c4 c5 c6), and c1, c2, c3, c4, c5, and c6 are the encoded information units.

[0057] The calculation method of LDPC encoding is: C(u) = (u1 u2 u3) × (100101; 010111; 001011), and the above formula can be abbreviated as C(u) = uG (where the addition between elements is modulo-two addition, i.e., "exclusive or"), where u is the original code and G is called the generation matrix. Among the encoded codewords, c1, c2, and c3 are equivalent to u1, u2, and u3; Transform the three expressions of c4, c5, and c6: Then it can be written in the form of matrix multiplication: (110100; 011010; 111001) × (c1 c2 c3 c4 c5 c6) T =(000) T , abbreviated as HC T (u) = 0, where H is called the parity-check matrix.

[0058] The decoding output is correct only when it satisfies the parity-check matrix / equation (i.e., the matrix multiplication can obtain a zero vector). Therefore, the receiving end can use H to determine whether there is an error in the transmission.

[0059] The present invention realizes fast interactive feedback based on the self-determination of the code pattern matrix characteristics, reducing the delay.

[0060] If the codeword after decoding can pass the decoding check, then send the correct feedback information Y to the sending end, indicating that the receiving end has completed the correct decoding of the received information. After outputting the decoded information, clear the buffer at the sending end and transmit a new encoded block; if it fails to pass the decoding check, then perform the following operations:

[0061] a. Perform a confidence threshold decision on the confidence feature of the soft information sequence output by LDPC decoding, cache the confidence features less than the threshold, and the quantity is D. The method for calculating the confidence feature is as follows: Use the magnitude of the absolute value of the soft information sequence of each information bit output after LDPC iterative decoding as the confidence feature, and the soft information sequence uses log-likelihood ratio information.

[0062] Specifically, the calculation of the log-likelihood ratio can be as follows: If the received signal is r, the ratio of the probability of correctly judging as 0 to the probability of correctly judging as 1 is the likelihood ratio, and taking the natural logarithm is the log-likelihood ratio LLR = ln[p(r|b = 0) / p(r|b = 1)].

[0063] In this embodiment, the confidence threshold threshold is a dynamic threshold (A) or a static threshold (B).

[0064] The dynamic threshold changes with the different signal-to-noise ratios SNR of the transmission, and the static threshold reduces the threshold decision complexity.

[0065] Among them, Rate is the code rate of the current LDPC code.

[0066] The present invention is based on the decoding confidence threshold, reduces the receiver buffer overhead and calculation complexity, and accelerates the LDPC decoding convergence through qualitative (decoding confidence) and quantitative (dual control of the interaction quantity and the confidence threshold), and improves the decoding accuracy.

[0067] b. Calculate the number of interactive information S and perform interactive information selection: If D >= S, then select the first S confidence feature corresponding bit positions Pos(S) in the buffer according to the ascending order of confidence as the directed information sequence flag for feedback; if D < S, then feedback all the D confidence feature corresponding bit positions Pos(D) in the buffer.

[0068] In this embodiment, the calculation method of the single - time interactive information number S is as follows:

[0069] It is determined by the algebraic relationship of the code rate Rate, the coding length Len, and the adaptive factor β of the current LDPC code:

[0070] S = Len×(0.028 - 0.02×Rate)×β (2)

[0071] The weight value can be appropriately adjusted according to the error - correction ability and broadband resource requirements. The adaptive factor β can be used to adjust the amount of interactive data, and can be balanced in terms of the number of interactions, error - correction ability, and broadband utilization according to system applications. For example, the larger β is, the larger the amount of data per interaction is, the stronger the error - correction ability is, the fewer the number of interactions is, and the less the interaction delay is, but the more broadband resources are occupied per single - time.

[0072] Through the dynamic adjustment of the confidence threshold and the adaptive factor, the present invention calculates the number of single interaction information S, improves the transmission quality, reduces the cache overhead, and realizes the dynamic balance of broadband resources and error correction capabilities, and achieves adaptive adjustment to meet different application requirements.

[0073] c, feedback the information N of decoding verification failure and the information identifier Pos(S) / Pos(D) of interaction. The sending end selects the corresponding interaction content from the buffer according to the Pos(S) / Pos(D) position.

[0074] S3. After the receiving end receives the interaction information from the sending end, update the soft information at the decoding end: update the soft information sequence LLR of the corresponding bit to the maximum value and decode again.

[0075] S4. Repeat the above steps until the decoding verification is satisfied or the maximum number of interactions is reached.

[0076] The adaptive LDPC encoding and decoding method with directional interaction of the present invention is based on the decoding convergence situation of LDPC codes and the characteristics of the code matrix. Under different communication conditions, different interaction contents are triggered through qualitative and quantitative selection of interaction information, realizing the adaptive adjustment of channel encoding and decoding. The present invention realizes the self-judgment feedback of decoding success or failure based on the verification method of the LDPC parity check matrix, reducing the delay. Based on the confidence characteristics of the decoding information convergence of LDPC codes, an adaptive strategy for updating the decoding decision probability is adopted to improve the decoding accuracy. The selection strategy of qualitative and quantitative information caching and dual control of interaction information reduces the overhead and the burden of the feedback link.

[0077] The present invention transmits the interaction and feedback information through a highly reliable transmission channel to ensure the reliability of the feedback and interaction information. A highly reliable transmission channel refers to a channel whose bit error rate meets the bit error rate requirement as a highly reliable transmission channel. For example, it can be determined according to the application needs. If the bit error rate requirement in the application can reach 1e10^-6, then a channel with a bit error rate <= 1e10^-6 is considered a highly reliable transmission channel.

[0078] Based on the convergence of LDPC iterative decoding at the receiving end, the confidence characteristics (decoding soft information) of each bit in the cache are determined based on the set decision threshold. The cache space size can be dynamically adjusted based on the channel condition (SNR) (A), or it can be based on a fixed threshold (B). The calculation of the interaction information adopts a dual-control mode. The buffer does not need to cache the eigenvalue of the entire information sequence that fails to pass the decoding verification. The cache space corresponds to the amount of data in a single interaction based on the threshold constraint. The threshold can be determined according to the actual application, and a suitable cache size can be opened with some spare space to be flexibly applicable to different applications. The amount of interaction data is jointly determined by the number of contents in the buffer under the threshold constraint and the number calculated by formula (2). In a low signal-to-noise ratio environment, the data cached with a dynamic threshold is less than that with a static threshold. In a high signal-to-noise ratio environment, the dynamic threshold value is much higher than the static threshold, and the amount of cached data will be more. Combining formula (2), different thresholds can be selected according to the application and the adaptive factor β in the reference formula can be adjusted. For example, the larger the β value, the more signal numbers in a single interaction, the fewer interaction times, the smaller the interaction delay, and the higher the throughput.

[0079] The present invention provides a transmitting end, which includes an encoding module, a caching module, and a directional information transmitting module; the encoding module receives the information to be encoded, performs LDPC encoding on the information sequence to be encoded and caches it in the caching module, and the encoding module transmits the LDPC encoding through a noisy channel to the receiving end; the caching module receives the feedback decoding verification failure information N and the interaction information identifier Pos fed back by the receiving end, selects the corresponding interaction content from the buffer according to the Pos position, and the interaction information identifier Pos is determined according to the adaptive LDPC encoding and decoding method of directional interaction of the present invention.

[0080] The transmitting end of the present invention receives the feedback decoding verification failure information N and the interaction information identifier Pos fed back by the receiving end, selects the corresponding interaction content from the buffer according to the Pos position, and triggers different interaction contents through qualitative and quantitative selection of interaction information under different communication conditions to realize the adaptive adjustment of channel encoding and decoding.

[0081] The present invention provides a receiving end, which includes a decoding module, a verification module, a decision buffer module, a directional interaction information calculation module, and a decoding information update module; the decoding module receives the encoded signal sent by the sending end, performs LDPC decoding and transmits it to the verification module and the decision buffer module; the verification module performs decoding verification. If the codeword after decoding can pass the decoding verification, it sends a correct feedback message Y to the sending end, indicating that the receiving end has completed the correct decoding of the received information; if it fails to pass the decoding verification, the directional interaction information calculation module performs the following operations: a. Perform a confidence threshold decision on the confidence feature of the soft information sequence output by LDPC decoding, and cache the confidence features less than the threshold, with the number being D; b. Calculate the number of interaction information S and perform interaction information selection: if D>=S, then select the first S confidence features corresponding to the bit positions Pos(S) in ascending order of confidence in the buffer as the directional information sequence flag for feedback; if D<S, then feedback all the bit positions Pos(D) corresponding to the D confidence features in the buffer; c. Feedback the decoding verification failure message N and the interaction information identifier Pos(S) / Pos(D) to the sending end; the decoding information update module receives the corresponding interaction content selected according to the Pos(S) / Pos(D) position sent by the sending end, updates the soft information at the decoding end, updates the LLR of the soft information sequence of the corresponding bit to the maximum value and transmits it to the decoding module for re-decoding.

[0082] Based on the decoding convergence situation of LDPC codes and the characteristics of the code matrix, the receiving end of the present invention triggers different interaction contents through qualitative and quantitative interaction information selection under different communication conditions, realizing the adaptive adjustment of channel coding and decoding. Based on the verification method of the LDPC parity-check matrix, the present invention realizes the self-judgment feedback of whether the decoding is successful, reducing the time delay. Based on the confidence feature of the decoding information convergence of LDPC codes, an adaptive strategy for updating the decoding decision probability is adopted to improve the decoding accuracy. The selection strategy of qualitative and quantitative information caching and dual control of interaction information reduces the overhead and the burden of the feedback link.

[0083] The present invention provides a coding and decoding system, which includes the sending end and the receiving end of the present invention.

[0084] Based on the decoding convergence situation of LDPC codes and the characteristics of the code matrix, the present invention triggers different interaction contents through qualitative and quantitative interaction information selection under different communication conditions, realizing the adaptive adjustment of channel coding and decoding. Based on the verification method of the LDPC parity-check matrix, the present invention realizes the self-judgment feedback of whether the decoding is successful, reducing the time delay. Based on the confidence feature of the decoding information convergence of LDPC codes, an adaptive strategy for updating the decoding decision probability is adopted to improve the decoding accuracy. The selection strategy of qualitative and quantitative information caching and dual control of interaction information reduces the overhead and the burden of the feedback link.

[0085] Based on the convergence situation and decoding characteristics of LDPC decoding, the present invention triggers a directional interaction mechanism. The interaction mechanism is triggered by verifying the success or failure of decoding based on the characteristics of the generator matrix and parity-check matrix at the LDPC encoding and decoding ends, reducing the processing delay; and the content of the interaction is selected according to the decoding convergence situation. Among them, the information content interaction and feedback are transmitted through a highly reliable channel to ensure the transmission quality of the feedback information between the transceiver ends; the number of interacted information is optimized and determined according to the code type and the channel conditions of the transmission, avoiding unnecessary consumption of transmission resources and caching overhead; the number of interactions is dynamically adjusted based on the bit error rate requirement of the service data and the channel resource overhead. The interacted information content is directionally selected based on the LDPC decoding convergence situation, and the soft information is updated at the decoding end to increase the decoding decision probability.

[0086] In the description of this specification, the descriptions with reference to the terms "preferred embodiment", "one embodiment", "some embodiments", "example", "specific example" or "some examples", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0087] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and purposes of the present invention. The scope of the present invention is defined by the claims and their equivalents.

Claims

1. An adaptive LDPC encoding and decoding method for directional interaction, characterized in that It includes the following steps: S1, the sending end performs LDPC encoding on the information sequence to be encoded and caches it; the sending end transmits the LDPC encoding through a noisy channel to the receiving end; S2, the receiving end receives the signal, performs LDPC decoding and LDPC decoding verification. If the decoded codeword can pass the decoding verification, the receiving end sends the correct feedback information Y to the sending end, indicating that the receiving end has completed the correct decoding of the received information. If it fails to pass the decoding verification, the following operations are performed: a, perform a confidence threshold decision on the confidence feature of the soft information sequence output by LDPC decoding, cache the confidence features less than the threshold, and the quantity is D; The method for calculating the confidence feature is: Take the magnitude of the absolute value of the soft information sequence of each information bit output after iterative decoding of the LDPC code as the confidence feature, and the soft information sequence uses log-likelihood ratio information; b, calculate the number of mutual information S and perform mutual information selection: if D > S or D = S, then select the first S confidence features corresponding bit positions Pos(S) in ascending order of confidence in the buffer as the directed information sequence flag for feedback; if D < S, then feedback all the bit positions Pos(D) corresponding to the D confidence features in the buffer; c, feedback the information N indicating the failure of decoding verification and the mutual information identifier Pos(S) / Pos(D) for interaction. The sending end selects the corresponding interaction content from the buffer according to the Pos(S) / Pos(D) position; S3, after the receiving end receives the interaction information from the sending end, update the soft information at the decoding end: update the soft information sequence LLR of the corresponding bit to the maximum value and decode again; S4, repeat the above steps until the decoding verification is satisfied or the maximum number of interactions is reached.

2. The adaptive LDPC encoding and decoding method for directional interaction according to claim 1, wherein The method for LDPC decoding verification is: u = (u0, u1, …, u k-1 ) is the information sequence to be encoded, G is the generator matrix of the LDPC code, H is the parity-check matrix of the LDPC code, and the codeword after encoding at the transmitter is c = u·G; Denote the codeword corresponding to the receiver after iterative decoding. After the LDPC decoding is completed, the codeword sequence and the parity-check matrix H satisfy the formula: c·H T = 0 (1) If the verification passes; otherwise the decoding fails.

3. The adaptive LDPC encoding and decoding method for directional interaction according to claim 1, wherein The confidence threshold threshold is a dynamic threshold (A) or a static threshold (B), The dynamic threshold varies with the signal-to-noise ratio SNR of the transmission, and the static threshold reduces the threshold decision complexity; Among them, Rate is the code rate of the current LDPC code.

4. The adaptive LDPC encoding and decoding method for directional interaction according to claim 1, characterized in that The calculation method for the number of mutual information S per interaction is: It is determined by the algebraic relationship of the code rate Rate, coding length Len of the current LDPC code, and the adaptive factor β: S = Len×(0.028 - 0.02×Rate)×β (2) 5. The adaptive LDPC encoding and decoding method for directional interaction according to claim 1, characterized in that The interaction and feedback information are transmitted through a highly reliable transmission channel.

6. A transmitting end, characterized in that, It includes an encoding module, a caching module, and a directed information sending module; The encoding module receives the information to be encoded, performs LDPC encoding on the information sequence to be encoded and caches it to the caching module, and the encoding module transmits the LDPC encoding through a noisy channel to the receiving end; The caching module receives the feedback decoding verification failure information N and the mutual information identifier Pos for interaction feedback from the receiving end, and selects the corresponding interaction content from the buffer according to the Pos position. The mutual information identifier Pos is determined according to the method described in any one of claims 1 to 5.

7. A receiving end, characterized in that, It includes a decoding module, a verification module, a decision caching module, a directed interaction information calculation module, and a decoding information update module; The decoding module receives the encoded signal sent by the sending end, performs LDPC decoding and transmits it to the verification module and the decision buffer module; The verification module performs decoding verification. If the decoded codeword can pass the decoding verification, it sends the correct feedback information Y to the sending end, indicating that the receiving end has completed the correct decoding of the received information. If it fails to pass the decoding verification, the directed mutual information calculation module performs the following operations: a. Perform a confidence threshold decision on the confidence feature of the soft information sequence output by LDPC decoding, cache the confidence features smaller than the threshold, and the quantity is D; b. Calculate the number of mutual information S and perform mutual information selection: if D > S or D = S, then select the first S confidence features corresponding bit positions Pos(S) in the buffer according to the ascending order of confidence as the directed information sequence flag for feedback; if D < S, then feedback all the bit positions Pos(D) corresponding to the D confidence features in the buffer; c. Feedback the decoding verification failure information N and the mutual information identifier Pos(S) / Pos(D) to the sending end; The decoding information update module receives the corresponding interactive content selected according to the Pos(S) / Pos(D) position sent by the sending end, updates the soft information at the decoding end, updates the LLR of the soft information sequence of the corresponding bit to the maximum value and transmits it to the decoding module for re-decoding.

8. A coding and decoding system, characterized in that, It includes the sending end described in claim 6 and the receiving end described in claim 7.

Images