Polarization code coding and decoding method and system of key bit secondary coding based on early stop criterion

Abstract

The invention discloses a polar code coding and decoding method and system based on key bit secondary coding of an early stop criterion. The method comprises the following steps: a coding end calculates the information bit number Kdynasc of each frame according to the total frame number and the code rate; the first Kdynamic reliable channels are taken as information bits, important data are distributed, the rest are frozen bits, other data are reserved, and n frames of coding strings are obtained through coding; information bits of each frame are divided into decoding key bits and conventional bits; encoding to obtain an (n + 1) th check frame; sending the first n frames of coding strings and the (n + 1) th frame of verification frame to a decoding end; decoding by a decoding end, and calculating a dynamic threshold rho of the current signal-to-noise ratio; calculating a partial path metric value gamma of the information bit, and reserving the maximum metric value as a bias item; calculating a relative partial path metric value epsilon of the information bit, if epsilon is less than rho, determining that decoding fails, and requesting to retransmit the coded frame, otherwise, continuing decoding; if all frame decoding results pass the CRC verification, the decoding is successful; and if any frame fails, activating the flag bit of the frame, and tracing back to the corresponding frame to decode again.

Description

Technical Field

[0001] The present invention belongs to the technical field of channel coding and communication, and in particular to a polar code coding and decoding method and system for secondary coding of key bits based on an early stopping criterion. Background Art

[0002] Polar code is a low-complexity channel coding proposed by Turkish professor E. Arikan in 2008. Professor Arikan concluded through rigorous mathematical derivation that polar code can achieve the channel capacity of binary discrete memoryless symmetric channel (B-DMC), which laid a solid foundation for the research of polar code. Because of its simple composition and low complexity, it is used as part of the control channel coding of new radio enhanced mobile broadband (EMBB) and ultra-reliable low latency communication (URLLC) in the fifth generation mobile communication technology (5G).

[0003] The coding algorithm of polar codes has attracted extensive attention from scholars, especially in the field of cascade coding. At present, the main goal of most cascade algorithms is often to improve channel capacity and reduce bit error rate. For example, the polarization product code algorithm, by cascading with the product code, first performs polarization coding and then product coding, can further reduce the system's bit error rate.

[0004] Since the decoding method of polar code is serial decoding, there is a large decoding delay. If a highly parallel belief propagation (BP) decoding algorithm is used, there is a high decoding complexity. Therefore, reducing the decoding complexity as much as possible while maintaining a low bit error rate has become a major problem at the decoding end. Correspondingly, there is also a lack of a coding method that can better balance complexity and bit error rate at the encoding end. Although the polarization product code algorithm effectively improves the decoding performance, the cascade of the two codes increases the decoding complexity. Chinese patent application number CN202310283865.2, application publication date June 23, 2023, discloses an invention patent application named "A polar code encoding method based on secondary encoding of key bits". Although this method has its advantages, it has a long decoding cycle in judging whether the current decoding is wrong and whether the encoded frame needs to be retransmitted. If effective judgment can be made, the bits destined to fail in decoding can theoretically be determined in advance through the early stopping criterion, so that the coded frame can be retransmitted earlier, further reducing the decoding complexity and delay. Therefore, in this field, there is an urgent need for a decoding technology with an early stopping criterion that can determine whether decoding has failed. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a polar code encoding and decoding method and system for secondary encoding of key bits based on an early stopping criterion.

[0006] The technical solution of the present invention is as follows:

[0007] A polar code encoding and decoding method for key bit secondary encoding based on an early stopping criterion includes a transmitting end (i.e., an encoding end) and a receiving end (i.e., a decoding end), and the specific steps are as follows:

[0008] S1. The encoder determines the number of transmission frames n and the single frame code length N according to the transmitted data, and calculates the number of information bits K of each frame according to the total number of frames n+1 and the code rate R dynamic ;

[0009] S2. Use the polarization weight construction method to obtain the channel reliability of each channel, and take the first K dynamic Reliable channels are used as information bits, and the remaining channels are used as frozen bits;

[0010] S3. Divide the data into important data and unimportant data, assign the important data to information bits, leave the unimportant data in frozen bits, perform polarization coding on the first n frames of polar codes, and obtain n frames of coded strings;

[0011] S4, according to whether the information bits are important for decoding, dividing the information bits of each of the first n frames into decoding key bits and decoding regular bits;

[0012] S5. Perform polar code encoding on all decoding key bit sets in the first n frames to obtain an (n+1)th frame check frame;

[0013] S6, the encoding end sends the first n frame encoding string and the (n+1)th frame check frame to the decoding end;

[0014] S7, the decoding end decodes the received coded data and calculates the dynamic threshold ρ under the current signal-to-noise ratio;

[0015] S8, calculating the partial path metric value γ corresponding to each information bit, and retaining the largest metric value as a bias term;

[0016] S9, during the decoding process, the relative partial path metric ε is calculated. If the metric corresponding to the current information bit is less than the dynamic threshold, the decoding fails and a request is made to retransmit the coded frame directly. Otherwise, the decoding of the next information bit is continued;

[0017] S10. If all frame decoding results pass the CRC check, the decoding is successful and the decoding is completed; if any frame CRC check fails, the flag bit of the frame is activated, and the corresponding frame is traced back to re-decode.

[0018] The present invention reduces the number of invalid traversals and reduces the complexity of the overall decoding algorithm.

[0019] As a preference, in step S1, the number of information bits K of each frame is recalculated dynamicThe formula is as follows:

[0020]

[0021] Among them, K inf represents the number of information bits in the initial polar code, Represents the floor function.

[0022] Preferably, the code length of each frame is consistent, and the code rate in the first n frames is consistent.

[0023] Preferably, CRC check bits are added to each frame.

[0024] Preferably, information bits with low reliability (i.e., reliability below a threshold) are used as key bits for decoding, and the remaining information bits are used as regular bits for decoding. The sum of all key bits of n frames plus a single-frame CRC shall not exceed the total code length of a single frame.

[0025] Preferably, the formula for calculating the dynamic threshold ρ under the current signal-to-noise ratio is as follows:

[0026] ρ=ln(FER / 1000)

[0027] Here, FER stands for frame error rate.

[0028] In by In the context of a set of independent and uniformly distributed random variables represented by , which correspond to the variables of the system, we can apply the Chernoff bound and Inequality to get the correct bit u i The probability that the relative partial path metric value is lower than the dynamic threshold ρ decreases exponentially as ρ decreases. The performance loss is set at 0.1×FER. Through Monte Carlo simulation experiments, it is found that for different polar code lengths N (256, 512, 1024), code rates R (1 / 3, 1 / 2, 2 / 3) and decoder list lengths L (1, 2, 4), in the low signal-to-noise ratio range, that is, FER = 0.1 to 0.001, the linear relationship between FER and ρ is as follows: Figure 3 In order to ensure that the performance loss of most codewords is within an acceptable range, the present invention obtains the above empirical formula by analyzing the corresponding data: ρ = ln(FER / 1000). When the signal-to-noise ratio SNR, the polar code length N, the code rate R and the decoder list length L are known, ρ can be directly calculated.

[0029] Preferably, in step S8, the formula for calculating the partial path metric value of the current decoding bit i is as follows:

[0030]

[0031] in, represents the path metric value, is the transmission probability of the transmitted signal. Take the maximum path metric value as the bias term B i .

[0032] Preferably, in step S9, the formula for calculating the relative partial path metric value of the current decoded bit i is as follows:

[0033]

[0034] in, represents the partial path metric, B i Represents the bias term.

[0035] Preferably, during the decoding process, the relative partial path metric corresponding to each decoded bit is calculated. If the current If it is less than the set threshold value ρ, the subsequent decoding is terminated immediately and the coded frame where the current bit is located is directly retransmitted. Otherwise, the current decoding process continues.

[0036] Preferably, in step S10, if the decoding results of the n coded frames all pass the CRC check, the decoding is considered successful, the decoding is completed, and the check frame is no longer decoded. If any frame fails to be decoded, the flag bit of the frame is activated. i =1, and decode the check frame for error detection.

[0037] As a preferred embodiment, the decoding process of the check frame is as follows: i = 0, the corresponding frame retains its result. i =1, the key bits are regarded as information bits and the check frame is decoded with traditional polar code. If the result does not pass the CRC check, the decoding fails and the decoding ends. Otherwise, the key bits are regarded as frozen bits through the backtracking operation of the check frame to decode the information frame again. If the CRC check is passed, the error correction is successful, otherwise the error correction fails.

[0038] The present invention also discloses a polar code encoding and decoding system for secondary encoding of key bits based on an early stopping criterion, which is used to execute the above method. The system includes an encoding end and a decoding end, wherein:

[0039] Encoding end: Determine the number of transmission frames n and the single frame code length N based on the transmitted data, and calculate the number of information bits K of each frame based on the total number of frames n+1 and the code rate R dynamic ;

[0040] The Gaussian construction method is used to obtain the channel reliability of each channel, and the first K dynamic Reliable channels are used as information bits, and the remaining channels are used as frozen bits;

[0041] The data is divided into important data and unimportant data, the important data is allocated to the information bit position, and the unimportant data is retained in the frozen bit position, and the polarization code of the first n frames is polarized to obtain an n-frame code string;

[0042] According to whether the information bits are important for decoding, the information bits of each frame in the first n frames are divided into decoding key bits and decoding regular bits;

[0043] Polar code all the decoding key bit sets in the first n frames to obtain the (n+1)th frame check frame;

[0044] Send the first n frames of coded string and the (n+1)th frame check frame to the decoding end;

[0045] Decoding end: decode the received coded data and calculate the dynamic threshold ρ under the current signal-to-noise ratio;

[0046] Calculate the partial path metric value γ corresponding to each information bit, and retain the largest metric value as the bias term;

[0047] During the decoding process, the relative partial path metric ε is calculated. If the metric ε corresponding to the current information bit is less than the dynamic threshold ρ, the decoding fails and the coded frame is directly requested to be retransmitted. Otherwise, the decoding of the next information bit is continued;

[0048] If all frame decoding results pass the CRC check, the decoding is successful and the decoding is completed; if any frame CRC check fails, the flag bit of the frame is activated and the corresponding frame is traced back for re-decoding.

[0049] Compared with the prior art, the advantages of the present invention are:

[0050] The polar code encoding and decoding method and system for key bit secondary encoding based on the early stopping criterion of the present invention calculates a dynamic threshold value under the current signal-to-noise ratio and a relative partial path metric value corresponding to each decoding bit during the decoding process at the decoding end. If the relative partial path metric value of the current decoding bit is less than the dynamic threshold value under the current signal-to-noise ratio, a decoding error is determined, and the information frame is directly retransmitted, thereby achieving the requirement of early stopping, achieving the judgment of the decoding result before the CRC check, and further reducing the complexity and delay of polar code decoding. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 The present invention is a flowchart of a polar code encoding and decoding method for secondary encoding of key bits based on an early stopping criterion in a preferred embodiment of the present invention.

[0052] Figure 2 The preferred embodiment of the present invention is that the encoding part is N=1024, K dynamic Structural diagram when =563, m=10.

[0053] Figure 3 The preferred embodiment of the present invention is for different N, R, L and SNR values, in P loss Display of the dynamic threshold ρ at ≤0.1×FER; each star point corresponds to a different quartet (N, R, L, SNR).

[0054] Figure 4 The preferred embodiment of the present invention is when N=1024, K dynamic =563, m=10, bit error rate simulation comparison with traditional polar code under different decoding methods.

[0055] Figure 5 The preferred embodiment of the present invention is when N=1024, K dynamic =563, m=10 and the simulation comparison with the traditional polar code complexity.

[0056] Figure 6 The present invention is a block diagram of a polar code encoding and decoding system for secondary encoding of key bits based on an early stopping criterion in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] like Figure 1 As shown, this embodiment proposes a polar code encoding and decoding method for key bit secondary encoding based on an early stopping criterion. Taking a polar code with a polar code length of 1024, a code rate of 0.5, and a number of transmitted frames of 10 as an example, this embodiment provides the following specific implementation steps:

[0059] Step 1: At the encoding end, the number of transmission frames is determined to be 10 and the single frame code length is 1024 according to the transmitted data, and the total number of frames is 11. The number of information bits of each frame is recalculated according to the total number of frames and the code rate:

[0060]

[0061] Among them, K inf represents the number of information bits in the initial polar code, Represents the floor function.

[0062] All information bits in 11 frames of polar codes with a code length of 1024 and a code rate of 0.5 are compressed into 10 frames. The number of polar codes in each frame is 563. At this time, the polar codes of each frame in the first 10 frames become polar codes with a code length of 1024 and a code rate of 0.55, but the total code rate of the 10 frames is 0.5.

[0063] Step 2: Use the polarization weight construction method to obtain the channel reliability of each channel, and take the first K dynamic reliable channels as information bits and the rest as frozen bits.

[0064] The specific process of the polarization weight construction method is as follows: Assuming that the polarization code with code length N corresponds to an index of each bit subchannel of 0≤j≤N-1, the binary expansion of j is obtained to obtain the expression {j n-1 , j n-2 , ..., j0), where n = log2N, j n-1 Is the highest bit, polarized channel The polarization weight PW is calculated as follows:

[0065]

[0066] Where β = 2 1 / 4 After obtaining all N polarization channels The PW values ​​are sorted from large to small. The higher the sorting, the more reliable the corresponding channel. The bit channel with the highest sorting is selected to transmit information. The following is an example of a polar code with a code length of 8:

[0067] The subchannel sequence is (W1, W2, W3, W4, W5, W6, W7, W8), and the corresponding subchannel bit index is (0, 1, 2, 3, 4, 5, 6, 7). Each subchannel bit index is binary expanded to obtain (000, 001, 010, 011, 100, 101, 110, 111). Finally, the binary expansion result is substituted into the PW calculation formula to obtain the polarization weight of each subchannel (0.000, 1.000, 1.189, 2.189, 1.414, 2.414, 2.603, 3.603). After sorting the PW values ​​in descending order, the channel reliability is ranked from high to low as (W8, W7, W6, W4, W5, W3, W2, W1).

[0068] Specifically in this embodiment, the sub-channels are sorted in ascending order according to the polarization weight, and the first 563 sub-channels are taken, that is, (W 128 , W 190 , ..., W 1022 , W 1023 ) is the information bit; the remaining 461 subchannels, namely (W1, W2, ..., W 1008 , W1024 ) is the frozen bit. i represents the i-th subchannel.

[0069] Step 3: Allocate important data to information bits and leave unimportant data in frozen bits. Perform polarization coding on the first 10 frames of polar codes to obtain 10 frames of coded strings.

[0070] Assume that the i-th frame coding sequence is u i , then the encoded sequence x i =u i ·G 10 Among them, G 10 is a 10th-order Kronecker product.

[0071] Step 4: According to whether the information bits are important for decoding, the information bits of each of the first n frames are divided into decoding key bits and decoding regular bits.

[0072] The information bits in the first 10 frames of the coded string are divided into decoding key bits and decoding regular bits according to the size of the polarization weight. Among them, the number of decoding key bits is 100, and the specific subchannel number is (W 915 , W 916 , ..., W 1022 , W 1023 ), the normal number of bits is 563-100=463, and the specific subchannel number is (W 128 , W 190 , ..., W 913 , W 914 ).

[0073] The decoding key bits of each frame in the first 10 frames are encoded and added to the check frame, and a 16-bit CRC code is added at the end. The sum of the total length of the key bits of the 10 frames plus the 16-bit CRC is required to be less than the total code length and less than the code length of 1024. On average, each frame can only be divided into The number of key bits.

[0074] Step 5: Perform polar code encoding on all decoding key bit sets in the first n frames to obtain the (n+1)th frame check frame.

[0075] The decoding key bits of each of the first 10 frames are encoded and added to the check frame, and a 16-bit CRC code is added at the end to form a check frame with a code length of 1024, 1016 information bits, and 8 frozen bits.

[0076] Step 6: The first n frames of coded strings and the (n+1)th frame check frame are sent from the encoding end to the decoding end.

[0077] The first 10 frames and the 11th frame are transmitted to the decoding end through the additive white Gaussian noise (AWGN) channel.

[0078] Step 7: The decoding end decodes the received coded data and calculates the dynamic threshold ρ under the signal-to-noise ratio.

[0079] Assuming the current SNR is 1.5 dB, N is 1024, and R is 0.5, the FER performance of the traditional polar code serial cancellation list SCL-4 (list size is 4) decoder at 1.5 dB is about 5×10 -2 , then the dynamic threshold ρ=ln(5×10 -2 / 1000)≈-9.90.

[0080] Step 8: Calculate the partial path metric value γ corresponding to each information bit, and retain the largest metric value as the bias term.

[0081] The decoding operation is performed on the first 10 frames of the coded string. In this embodiment, only the decoding of 8 information bits in a single frame is demonstrated. If the signal received by the decoder is: By formula Obtain the channel log-likelihood ratio LLR value:

[0082]

[0083] According to the formula in, It represents the LLR value corresponding to the jth decoded bit in the total code length N. After calculation, the path metric PM value of the channel is as follows:

[0084]

[0085] Finally, substitute the formula in, represents the path metric value, The partial path metric for each decoded bit is obtained as the transmission probability of the transmitted signal as follows:

[0086] (0.04, 1.57, 6.18, 9.42, 12.73, 12.73, 12.73, 12.73)

[0087] Take the maximum path metric value of all decoded bits as the bias term B i , then B i =12.73.

[0088] Step 9: During the decoding process, the relative partial path metric ε is calculated. If the metric corresponding to the current information bit is less than the dynamic threshold, the decoding is considered to have failed and a request is made to retransmit the coded frame directly. Otherwise, the decoding of the next information bit is continued.

[0089] According to the formula The relative partial path metric of the current decoded bit i is calculated as follows:

[0090]

[0091] After comparison, if the relative partial path metric values ​​of the first and second information bits are less than the set dynamic threshold ρ, it is considered that the current bit decoding fails and the coded frame is directly retransmitted.

[0092] Step 10: If all frame decoding results pass the CRC check, the decoding is considered successful and the decoding is completed. If any frame fails, the flag bit corresponding to the frame in the decoded frame is activated, and the decoded check frame is used for error correction.

[0093] If the decoding fails, the flag bit of the frame in the check frame will be activated. i =1, for f i = 0, the corresponding frame retains its result. i =1, the key bits are regarded as information bits and the check frame is decoded with traditional polar code. If the result does not pass the CRC check, the decoding fails and the decoding ends. Otherwise, the key bits are regarded as frozen bits through the backtracking operation of the check frame to decode the information frame again. If the CRC check is passed, the error correction is successful. Otherwise, the error correction fails and the coded frame is requested to be retransmitted.

[0094] The following is combined with Figure 4 , 5 The superiority of the polar code encoding and decoding method of the key secondary encoding based on the early stopping criterion in this embodiment is illustrated.

[0095] Figure 4 The preferred embodiment of the present invention is a polar code decoding method based on the key secondary coding of the early stopping criterion in N = 1024, K dynamic =563, m=10 and the bit error rate simulation comparison diagram of the traditional polar code under different decoding modes. The curve with triangles and diamonds on the solid line represents the decoding method proposed by the present invention. Figure 4 It can be seen that, under the same decoding list length L, the method of the present invention has almost no loss in decoding performance compared with the key bit secondary encoding method using traditional SCL decoding. Figure 5 It is a complexity comparison chart. The curve with triangles and diamonds on the solid line represents the decoding method proposed in the present invention. It can be seen that under medium and low signal-to-noise ratios and under the same decoding list length L, the decoding complexity of the method of the present invention is significantly reduced compared with the traditional SCL decoding method. The complexity reduction is about 9.3% at 1.5dB.

[0096] like Figure 6As shown, this embodiment discloses a polar code encoding and decoding system for key bit secondary encoding based on an early stopping criterion, which is used to execute the above method. The system includes an encoding end and a decoding end, wherein:

[0097] Encoding end: Determine the number of transmission frames n and the single frame code length N based on the transmitted data, and calculate the number of information bits K of each frame based on the total number of frames n+1 and the code rate R dynamic ;

[0098] The channel reliability of each channel is obtained by using the polarization weight construction method, and the first K dynamic Reliable channels are used as information bits, and the remaining channels are used as frozen bits;

[0099] The data is divided into important data and unimportant data, the important data is allocated to the information bit position, and the unimportant data is retained in the frozen bit position, and the polarization code of the first n frames is polarized to obtain an n-frame code string;

[0100] According to whether the information bits are important for decoding, the information bits of each frame in the first n frames are divided into decoding key bits and decoding regular bits;

[0101] Polar code all the decoding key bit sets in the first n frames to obtain the (n+1)th frame check frame;

[0102] Send the first n frames of coded string and the (n+1)th frame check frame to the decoding end;

[0103] Decoding end: decode the received coded data and calculate the dynamic threshold ρ under the current signal-to-noise ratio;

[0104] Calculate the partial path metric value γ corresponding to each information bit, and retain the largest metric value as the bias term;

[0105] During the decoding process, the relative partial path metric ε is calculated. If the metric ε corresponding to the current information bit is less than the dynamic threshold ρ, the decoding fails and the coded frame is directly requested to be retransmitted. Otherwise, the decoding of the next information bit is continued;

[0106] If all frame decoding results pass the CRC check, the decoding is successful and the decoding is completed; if any frame CRC check fails, the flag bit of the frame is activated and the corresponding frame is traced back for re-decoding.

[0107] For other contents of this embodiment, please refer to the above method embodiment.

[0108] The above embodiments further illustrate the purpose, technical solutions and advantages of the present invention in detail. It should be understood that the above embodiments are only preferred implementation modes of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made to the present invention within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A polar code encoding and decoding method for key bit secondary encoding based on early stopping criterion, characterized in that: The steps include: S1. The encoder determines the number of transmission frames n and the single frame code length N according to the transmitted data, and calculates the number of information bits K of each frame according to the total number of frames n+1 and the code rate R dynamic ; S2. Use the polarization weight construction method to obtain the channel reliability of each channel, and take the first K dynamic Reliable channels are used as information bits, and the remaining channels are used as frozen bits; S3. Divide the data into important data and unimportant data, assign the important data to information bits, keep the unimportant data in frozen bits, perform polarization coding on the first n frames of polar codes, and obtain n frames of coded strings; S4, according to whether the information bits are important for decoding, dividing the information bits of each frame in the first n frames into decoding key bits and decoding regular bits; S5. Perform polar code encoding on all decoding key bit sets in the first n frames to obtain an (n+1)th frame check frame; S6, the encoding end sends the first n frame encoding string and the (n+1)th frame check frame to the decoding end; S7, the decoding end decodes the received coded data and calculates the dynamic threshold ρ under the current signal-to-noise ratio; S8, calculating the partial path metric value γ corresponding to each information bit, and retaining the largest metric value as a bias term; S9, during the decoding process, the relative partial path metric ε is calculated. If the metric ε corresponding to the current information bit is less than the dynamic threshold ρ, the decoding fails and a request is made to retransmit the coded frame directly. Otherwise, the decoding of the next information bit is continued. S10. If all frame decoding results pass the CRC check, the decoding is successful and the decoding is completed; if any frame CRC check fails, the flag bit of the frame is activated, and the corresponding frame is traced back for re-decoding.

2. The polar code encoding and decoding method for key bit secondary encoding based on early stopping criterion according to claim 1, characterized in that: In step S1, the number of information bits K of each frame is recalculated dynamic The formula is as follows: Among them, K inf represents the number of information bits in the initial polar code, Represents the floor function.

3. The polar code encoding and decoding method of key bit secondary encoding based on early stopping criterion according to claim 1, characterized in that: In step S1, the code length in each frame is consistent, and the code rate in the first n frames is consistent; Alternatively, in step S3 and step S5, CRC check bits are added to each frame.

4. The polar code encoding and decoding method for key bit secondary encoding based on early stopping criterion according to claim 1, characterized in that: In step S4, information bits with reliability lower than the threshold are used as decoding key bits, and the remaining information bits are used as decoding regular bits. The sum of all key bits of n frames plus the CRC check bits of a single frame does not exceed the total code length of a single frame.

5. The polar code encoding and decoding method of key bit secondary encoding based on early stopping criterion according to claim 1, characterized in that: In step S7, the formula for calculating the dynamic threshold ρ under the current signal-to-noise ratio is as follows: ρ=ln(FER / 1000) Here, FER stands for frame error rate.

6. The polar code encoding and decoding method of key bit secondary encoding based on early stopping criterion according to claim 1, characterized in that: In step S8, the formula for calculating the partial path metric value of the current decoding bit i is as follows: in, represents the path metric value, is the transmission probability of the transmitted signal; the maximum path metric is taken as the bias term B i .

7. The polar code encoding and decoding method of key bit secondary encoding based on early stopping criterion according to claim 6, characterized in that: In step S9, the formula for calculating the relative partial path metric value of the current decoded bit i is as follows: in, represents the partial path metric, B i represents the bias term; During the decoding process, the relative partial path metric corresponding to each decoded bit is calculated If the current If it is less than the set threshold ρ, the subsequent decoding is terminated immediately and the coded frame where the current bit is located is directly retransmitted; otherwise, the current decoding process continues.

8. The polar code encoding and decoding method for key bit secondary encoding based on early stopping criterion according to any one of claims 1 to 7, characterized in that: In step S10, if the decoding results of the n coded frames all pass the CRC check, the decoding is considered successful, the decoding is completed, and the check frame is no longer decoded; If any frame decoding CRC check fails, the flag bit of the frame is activated. i =1, and the decoding check frame is error checked.

9. The polar code encoding and decoding method of key bit secondary encoding based on early stopping criterion according to claim 8, characterized in that: The decoding process of the check frame is as follows: i = 0, the corresponding frame retains its result. i =1, the key bits of the corresponding frame are regarded as information bits, and the traditional polar code decoding is performed on the check frame. If the result does not pass the CRC check, the decoding fails and the decoding ends. Otherwise, the key bits are used to be regarded as frozen bits through the backtracking operation of the check frame to decode the information frame again. If the CRC check is passed, the error correction is successful, otherwise the error correction fails and the decoding ends.

10. A polar code encoding and decoding system with key bit secondary encoding based on early stopping criterion, used to execute the method according to any one of claims 1 to 9, characterized in that: The system comprises an encoding end and a decoding end, wherein: Encoding end: Determine the number of transmission frames n and the single frame code length N based on the transmitted data, and calculate the number of information bits K of each frame based on the total number of frames n+1 and the code rate R dynamic ; The channel reliability of each channel is obtained by using the polarization weight construction method, and the first K dynamic Reliable channels are used as information bits, and the remaining channels are used as frozen bits; The data is divided into important data and unimportant data, the important data is allocated to the information bit position, and the unimportant data is retained in the frozen bit position, and the polarization code of the first n frames is polarized to obtain an n-frame code string; According to whether the information bits are important for decoding, the information bits of each frame in the first n frames are divided into decoding key bits and decoding regular bits; Polar code all the decoding key bit sets in the first n frames to obtain the (n+1)th frame check frame; Send the first n frames of coded string and the (n+1)th frame check frame to the decoding end; Decoding end: decode the received coded data and calculate the dynamic threshold ρ under the current signal-to-noise ratio; Calculate the partial path metric value γ corresponding to each information bit, and retain the largest metric value as the bias term; During the decoding process, the relative partial path metric ε is calculated. If the metric ε corresponding to the current information bit is less than the dynamic threshold ρ, the decoding fails and the coded frame is directly requested to be retransmitted. Otherwise, the decoding of the next information bit is continued; If all frame decoding results pass the CRC check, the decoding is successful and the decoding is completed; if any frame CRC check fails, the flag bit of the frame is activated and the corresponding frame is traced back for re-decoding.

Images