A self-synchronous decoding method based on polar codes

Abstract

The present application relates to the field of communication technology, specifically to a self-synchronous decoding method based on polar code, comprising: selecting part of the frozen bit position to embed a preset synchronization identification sequence when constructing the polar code encoding structure at the sending end, introducing synchronization auxiliary capability under the premise of maintaining non-information bearing attribute; based on the frozen bit symbol stability change trend and path metric value mutation characteristics, identifying multiple synchronization candidate starting positions during the bit stream receiving process at the receiving end; performing restricted depth decoding on each candidate point, calculating the Hamming consistency score of the frozen bit and the synchronization sequence, and constructing a synchronization confidence function by weighting fusion with the path metric value to determine the optimal synchronization starting point and then perform the complete polar code decoding process to recover the original information bit. The method does not require an independent synchronization field, has high synchronization accuracy and decoding efficiency, and is suitable for low latency and low resource occupation scenarios in high-speed communication.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to a self-synchronous decoding method based on polar code. BACKGROUND

[0002] With the development of high-performance communication systems such as 5G and Internet of Things, higher requirements are put forward for the decoding efficiency and synchronization accuracy of channel coding methods. Polar code, as a channel coding scheme adopted by 5G standard, has attracted widespread attention due to its capacity approximation performance and structural programmability.

[0003] Existing polar code communication systems rely on dedicated synchronization fields such as preambles and frame headers for frame synchronization. Such methods not only occupy channel resources and reduce effective code rate, but also are susceptible to noise interference, leading to synchronization errors. In addition, the traditional synchronization mechanism is separate from the decoding process, and cannot dynamically adjust the synchronization judgment based on actual decoding path feedback information, resulting in increased misjudgment rate and increased synchronization delay, which is not suitable for low-latency synchronization requirements in high-speed dynamic scenarios. Therefore, there is an urgent need for a mechanism that can integrate path behavior and achieve synchronization and decoding simultaneously without relying on external synchronization sequences, in order to improve the efficiency and robustness of the overall communication system.

[0004] Therefore, the present application provides a self-synchronous decoding method based on polar code. SUMMARY

[0005] The present application provides a self-synchronous decoding method based on polar code, which aims to provide a polar code self-synchronous decoding method that does not require additional synchronization fields, can identify the starting position of the frame in real time during the receiving process and complete information restoration, in order to reduce channel resource overhead and improve synchronization accuracy and decoding efficiency.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] The present application provides a self-synchronous decoding method based on polar code, which includes:

[0008] When constructing the polar code encoding structure at the sending end, a part of the frozen bit positions is selected to embed a preset synchronization identification sequence, and the frozen bits are set to predefined non-zero values on the premise of maintaining the non-information bearing role, which is used for synchronization auxiliary identification at the decoding end;

[0009] During the process of receiving the polar code encoding bit stream at the receiving end, based on the symbol stability change trend of the frozen bits in the path expansion and the mutation characteristics of the path metric value, a plurality of synchronization candidate starting positions are identified in real time;

[0010] For each candidate position, a restricted path decoding operation is performed, the frozen bit restoration value is extracted and compared with the preset synchronization identification sequence to obtain a consistency score;

[0011] The path metric value is combined with the consistency score to construct a synchronization confidence function, the position with the highest score is selected as the synchronization starting point, and the complete polar code decoding process is performed starting from the synchronization starting point to output the restored information bit sequence.

[0012] As a preferred technical solution of the application, the synchronization identification sequence is composed of a binary sequence of fixed length, which is generated by a linear feedback shift register and embedded in selected frozen bit positions, and is distributed in a bit interval manner to avoid significant interference with the overall structure performance of the polar code.

[0013] As a preferred technical solution of the application, the position selection of the frozen bits includes the following steps:

[0014] According to the code length parameter of the polar code, the reliability of each bit channel is estimated by using the Bartlett parameter method, and a bit channel reliability ranking list is generated;

[0015] According to the set code rate, the bits with high reliability ranking are selected as information bits from the ranking list, and the remaining positions are defined as a frozen bit set;

[0016] In the frozen bit set, the frozen bits with reliability parameter values not lower than a set value are selected to embed the synchronization identification sequence, taking into account the structural integrity and decoding recognizability, and are distributed in a fixed interval manner;

[0017] The bit values of the preset synchronization identification sequence are embedded into the embedding positions, so that the part of frozen bits have synchronization auxiliary function without carrying valid information.

[0018] As a preferred technical solution of the application, the symbol stability change trend includes:

[0019] The decoding values of the frozen bits in the decoding path are continuously monitored at the receiving end, and in each sliding window, the symbol consistency change of the frozen bit positions in the path list within the window range is counted;

[0020] If the symbol consistency of the frozen bits in a sliding window significantly decreases, which is manifested as significant inconsistency in the decoding values of the majority of frozen bits among the candidate paths, and this phenomenon lasts for more than a set amplitude threshold, it is determined that the starting position of the current window is a synchronization candidate starting point.

[0021] As a preferred technical solution of the application, the mutation behavior of the path metric value includes:

[0022] In the decoding process, the path metric change trend of each candidate path is continuously monitored, when the path average metric value in the path set corresponding to the current time changes by more than a set change threshold relative to the last time, or the number of effective paths retained in the path expansion process decreases by more than a set proportion in a single bit decision, the position corresponding to the decoding time is determined as a synchronization candidate starting position.

[0023] As a preferred technical solution of the application, the consistency comparison includes the following steps:

[0024] For each candidate starting position, perform a partial path decoding operation of the polar code to obtain the restored value of the frozen bit under the current path;

[0025] Compare the frozen bit restoration value with the synchronization identification sequence embedded by the sending end bit by bit, and calculate the Hamming distance as a consistency score;

[0026] The smaller the Hamming distance, the higher the matching degree, and the more likely the corresponding candidate starting position is the real synchronization starting point.

[0027] As a preferred technical solution of the application, the synchronization confidence function includes:

[0028] Based on the fusion of path metric information and frozen bit consistency score;

[0029] Among them, the path metric information is used to reflect the overall confidence degree of the decoding path, and a relative confidence ranking is formed among different candidate paths;

[0030] The consistency score is used to measure the consistency degree of the frozen bit in the decoding path with the preset synchronization identification sequence, and the consistency score is obtained by performing a restricted depth path decoding on each candidate synchronization starting position;

[0031] The above two types of information are fused according to a preset weighting strategy to form a comprehensive score index for synchronization starting point determination;

[0032] The comprehensive score result is used to sort a plurality of candidate synchronization starting positions, and the position with the highest score is selected as the final synchronization starting point.

[0033] As a preferred technical solution of the application, the complete polar code decoding process includes:

[0034] The SCL method is adopted, and the correctness of a plurality of candidate decoding paths is judged in combination with the CRC;

[0035] The validity verification step of the synchronization starting point comprises: after completing the polar code list decoding at the selected synchronization starting position, if there is at least one path passing the CRC check in the set of decoding paths, it is confirmed that the current synchronization starting point is a valid synchronization point, and the corresponding decoding result is output;

[0036] If all paths do not pass the CRC check, the current synchronization starting point is abandoned, the next candidate synchronization position is selected in the order of synchronization confidence scores, and the complete decoding and checking process is re-executed;

[0037] When all candidate synchronization positions do not pass the CRC check, it is determined that the current data block synchronization fails, and a system-level resynchronization mechanism is triggered or waiting for the next identifiable synchronization block.

[0038] The beneficial effects of the present application are:

[0039] 1. The synchronization mechanism described in the present application fully excavates the polar code decoding path information and does not rely on additional synchronization frame headers or pilot code structures. Through the synchronization reconstruction of frozen bits, path statistical analysis and score fusion mechanism, the self-synchronization process can be executed in cooperation with the standard SCL decoding process, which is suitable for the existing hardware platform structure. Especially in the communication environment where there is multi-path interference, frame boundary drift or uncertain receiving state, the method has higher robustness and synchronization flexibility, and significantly improves the overall anti-out-of-step ability and decoding success rate of the system.

[0040] 2. The synchronization starting point selection in the present application is based on the weighted fusion of the Hamming consistency score of the frozen bits and the path metric value, which takes into account the local sequence matching accuracy and the overall path reliability. Compared with the synchronization method of single structure or single criterion, it has higher accuracy, especially suitable for high-speed communication applications with strict error rate control. BRIEF DESCRIPTION OF DRAWINGS

[0041] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation of the present application. In the drawings:

[0042] Figure 1 is a flowchart of a self-synchronization decoding method based on polar codes provided by the present application; DETAILED DESCRIPTION

[0043] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and do not limit the present application.

[0044] Embodiment one:

[0045] A self-synchronization decoding method based on polar codes, as shown in Figure 1 , comprises:

[0046] In the process of constructing the polar code encoding structure at the sending end, a part of the frozen bit positions are selected to embed a preset synchronization identification sequence, the frozen bits are set to a predefined non-zero value on the premise of maintaining a non-information bearing role, and are used for synchronization auxiliary identification at the decoding end;

[0047] Further, the synchronization identification sequence is composed of a binary sequence of fixed length, which is generated by a linear feedback shift register and embedded in the selected frozen bit positions. When embedded, the bit interval method is used for distribution to avoid significant interference with the overall structure performance of the polar code.

[0048] The selection of the position of the frozen bit includes the following steps:

[0049] According to the code length parameter of the polar code, the reliability of each bit channel is estimated by using the Bhattacharyya parameter method, and a bit channel reliability sorting list is generated;

[0050] According to the set code rate, the bit positions with Bhattacharyya parameter values lower than a first preset threshold value are selected from the sorting list as information bits, and the remaining positions are selected as a frozen bit set;

[0051] In the frozen bit set, a number of frozen bits with Bhattacharyya parameter values higher than a second preset threshold value are selected as embedding positions of the synchronization identification sequence, and are uniformly distributed in the code word in a fixed interval manner;

[0052] The bit values of the preset synchronization identification sequence are embedded into the embedding positions, so that the part of the frozen bits have a synchronization auxiliary function without bearing valid information.

[0053] Specifically, first, the code length parameter of the polar code is set , that is, the total number of encoding bits, and the target information bit number is set , so as to determine the total number of frozen bits as .

[0054] Next, the reliability of each polar bit channel with a length of is estimated one by one by using the Bhattacharyya parameter method, and an ordered list of bit channel reliability values is generated. Based on the channel transmission function, the Bhattacharyya upper bound index of each bit channel is obtained by recursive calculation, and the smaller the value is, the more reliable the corresponding bit channel is.

[0055] According to the reliability value sorting result, the first most reliable bit positions are selected as information bits, and the remaining bit positions are defined as frozen bit positions. These frozen bit positions are recorded as a set .

[0056] In the frozen bit set In the middle, again from low to high, the order of the bar value, from which a number of the lowest reliability bit positions are selected as the embedding position of the synchronization identification sequence. The length of the synchronization identification sequence is set to , then take from the frozen bit set as the embedding site, recorded as the set , which satisfies .

[0057] In order to reduce the disturbance of embedding on the performance of polar code structure, the above embedding site is selected in a fixed bit interval manner. For example, in the frozen bit set , select from the end to the front in steps , so that the embedding position is evenly distributed in the entire code word. The embedding step can be calculated and determined according to the code length and the identification sequence length , which satisfies the approximate uniform coverage.

[0058] Then, using a linear feedback shift register (LFSR) generator, a set of fixed synchronization identification sequences is generated as a binary sequence vector . The feedback polynomial and initial state of the LFSR are pre-shared by both parties to ensure that the synchronization sequence can be restored at the receiving end.

[0059] Finally, in the process of constructing the polar code, each bit value of the above generated synchronization identification sequence is embedded in the corresponding frozen site ; the embedding process keeps the bits on other frozen sites still set to all zeros, thereby ensuring that the structure of the entire polar code still satisfies the basic design constraints.

[0060] Through the above steps, the sending end completes the synchronization auxiliary function of part of the frozen bits in the polar code coding structure, so that it can mark the synchronization position without carrying information bits, which facilitates the receiving end to achieve accurate synchronization identification without auxiliary synchronization sequence.

[0061] In the process of receiving the polar code coding bit stream at the receiving end, based on the symbol stability change trend and the mutation characteristics of the path metric value of the frozen bits in the path expansion, a plurality of synchronization candidate starting positions are identified in real time;

[0062] Further, the symbol stability change trend includes:

[0063] The decoding values of the frozen bits in the decoding path are continuously monitored at the receiving end, and in each sliding window, the symbol consistency change of the frozen bit positions in the path list within the window range is counted.

[0064] If the consistency of frozen bit symbols in a sliding window significantly decreases, which is manifested as the decoding values of most frozen bits in the candidate paths are significantly inconsistent, and the phenomenon lasts more than a set amplitude threshold, it is determined that the starting position of the current window is a synchronization candidate starting point.

[0065] The mutation behavior of the path metric value includes:

[0066] In the decoding process, the path metric change trend of each candidate path is continuously monitored. When the path average metric value in the path set corresponding to the current time changes by more than a set change threshold relative to the last time, or the number of effective paths retained in the path expansion process decreases by more than a set proportion in a single bit decision, the position corresponding to the decoding time is determined as a synchronization candidate starting position.

[0067] Specifically, the length of the sliding window is set to , the size of the path list is , that is, the number of retained decoding paths is , the frozen bit set is , which contains a total of frozen bit positions.

[0068] In the process of bit stream arriving one by one, the receiving end uses the sequential scanning method to perform path expansion. After receiving a new bit, a new path set is constructed. For the time when the th bit is received, a sliding window with a length of is backtracked from the current position, that is, all bit index sets belonging to the frozen bit set in the position interval are investigated, denoted as .

[0069] For each frozen bit position in the sliding window, the decoding value of the position on the retained decoding paths is extracted , and the consistency degree of the position is counted. The specific method is as follows:

[0070] Let the bit value appearing most frequently on the frozen bit be , and the symbol consistency score of the position is defined as:

[0071] ;

[0072] Where is an indicator function, which takes 1 if the bracket is true, and 0 otherwise.

[0073] The consistency scores of all frozen bit positions in the sliding window are averaged to obtain the consistency index of the current window:

[0074] ;

[0075] Continuously record the average consistency index of multiple sliding windows. If the current window Compared to the previous window The reduction exceeds the set threshold , that is, satisfying:

[0076] ;

[0077] The current window starting position Marks a potential synchronization candidate starting position.

[0078] The path metric is defined in the form of log-likelihood cost under the polar code serial cancellation list decoding path. Specifically, for any path , in the judgment bits, let the corresponding log-likelihood ratio be , the path metric update formula is as follows:

[0079] ;

[0080] in For the path in bits The decision value at is the log-likelihood ratio of the bit, ensuring that the path update is consistent with the decoding bias information direction.

[0081] At the same time, monitor the path metric mean of the decoding path. Bit position, corresponding to The path metric mean of the first path is There are significant mutations at the following positions:

[0082] ;

[0083] in is the average path metric value of the current location, is the mutation threshold.

[0084] If the consistency of the frozen bit symbol drops by more than And the path metric mutation amplitude exceeds , then the position is determined as a valid synchronization candidate starting point, recorded and used for subsequent partial path decoding and synchronization scoring operations.

[0085] For each candidate position, a restricted path decoding operation is performed to extract the frozen bit restoration value and compare it with the preset synchronization recognition sequence to obtain a consistency score;

[0086] Further, the consistency comparison includes the following steps:

[0087] For each candidate starting position, a polar code partial path decoding operation is performed to obtain the restored value of the frozen bit under the current path;

[0088] The frozen bit restored value is bitwise compared with the synchronization identification sequence embedded by the sending end, and the Hamming distance is calculated as the consistency score;

[0089] The smaller the Hamming distance, the higher the matching degree, and the more likely the corresponding candidate starting position is the real synchronization starting point.

[0090] Specifically, the limited decoding depth is set as , indicating that only the first bit is decoded in the bit stream starting at each candidate starting position.

[0091] The polar code path expansion decoding is started at the candidate starting position , and a simplified SCL decoder is used to expand bit. In this process, the bit restored value of the corresponding frozen bit position set in the decoding path is extracted to form a frozen bit restored sequence:

[0092] ;

[0093] The synchronization identification sequence is a binary sequence preset by the sending end and embedded in the frozen bit position, denoted as:

[0094] ;

[0095] wherein , indicates the number of frozen bits for synchronization identification.

[0096] The frozen bit restored value in the path decoding result is bitwise compared with the preset synchronization identification sequence , and the Hamming distance is used to measure the number of inconsistent bits:

[0097] ;

[0098] wherein is an indicator function, which takes 1 if the bracket is true, and 0 otherwise.

[0099] The Hamming distance is normalized and inversely mapped into a consistency score :

[0100] ;

[0101] The closer the consistency score is to 1, the higher the matching degree of the frozen bit restoration value and the synchronization identification sequence, and the greater the possibility that the candidate starting point is the real synchronization starting point.

[0102] The scores of all candidate positions are fused Output, for subsequent construction of synchronization confidence function and selection of final synchronization starting point.

[0103] The path metric value and the consistency score are weighted and fused to construct a synchronization confidence function, and the position with the highest score is selected as the synchronization starting point. The complete polar code decoding process is performed starting from the synchronization starting point, and the restored information bit sequence is output.

[0104] Further, the synchronization confidence function comprises:

[0105] Based on the fusion of path metric information and frozen bit consistency score;

[0106] Wherein, the path metric information is used to reflect the overall confidence degree of the decoding path, and the relative confidence ranking is formed among different candidate paths;

[0107] The consistency score is used to measure the consistency degree of the frozen bit in the decoding path and the preset synchronization identification sequence, and the consistency score is obtained by performing restricted depth path decoding on each candidate synchronization starting position;

[0108] The above two types of information are fused according to the preset weighting strategy to form a comprehensive score index for synchronization starting point determination;

[0109] The comprehensive score result is used to sort a plurality of candidate synchronization starting positions, and the position with the highest score is selected as the final synchronization starting point.

[0110] Specifically, for each candidate synchronization starting position , the path with the highest score in the restricted depth decoding is selected , and the path metric value of the corresponding path at depth is extracted. Since the path metric value is in the form of negative log likelihood accumulation, it needs to be normalized and processed in reverse. Let the maximum metric value of all candidate paths be , and the minimum be , then for each candidate position, calculate:

[0111] ;

[0112] Wherein , the larger the value, the better the path.

[0113] To ensure that the fused score has consistent physical meaning, the path confidence degree after reverse normalization Consistency score with frozen bits Perform weighted fusion to obtain synchronization confidence score:

[0114] ;

[0115] in is the preset fusion weight coefficient.

[0116] Score set for all candidate points Sort and get the highest-scoring candidate synchronization point index:

[0117] ;

[0118] Determine its corresponding position The final synchronization starting point.

[0119] Furthermore, the SCL method is used to Starting with , perform full polar code decoding on the received code stream, using a list width of SCL decoder; and combined with the CRC check mechanism to determine the validity of each decoding path;

[0120] The steps for verifying the validity of the synchronization starting point include:

[0121] After completing polar code decoding at the selected synchronization start position, if there is at least one path in the decoding path set that passes the CRC check, the current synchronization start point is confirmed to be a valid synchronization point, and the corresponding decoding result is output;

[0122] If all paths fail the CRC check, the current synchronization starting point is abandoned, and the next candidate synchronization position is selected in the order of synchronization confidence score, and the complete decoding and verification process is performed again;

[0123] When all candidate synchronization positions fail the CRC check, it is determined that the current data block synchronization has failed and the system waits for the next identifiable synchronization block.

[0124] Through the above-mentioned fusion mechanism and the complete decoding path verification process, reliable synchronization starting point determination and decoding result recovery can be achieved without the help of additional synchronization sequences, thereby improving the overall system anti-interference ability and communication efficiency.

[0125] Example 2:

[0126] In order to verify the advantages of the self-synchronous decoding method of the present invention in significantly improving synchronization accuracy and system efficiency without introducing additional synchronization sequences, this embodiment uses polar code , number of information bits Based on this, a complete self-synchronous decoding process is described, from the construction at the sending end to the synchronous identification and decoding output at the receiving end. The specific steps are as follows:

[0127] Constructing a frozen set after sorting using Bhattacharyya parameters , from which 16 frozen bits with reliability parameter values ​​not lower than the set value range are selected to form a synchronous recognition embedding set .

[0128] Using LFSR to generate synchronization recognition sequence , in the frozen position The remaining frozen bits are still set to 0.

[0129] After the bit stream is received, the candidate synchronization starting point is scanned with a sliding window with a step size of 1 , and map the frozen bit positions according to the polar code construction rules.

[0130] At each candidate starting point, the execution depth is Path extension, only get the frozen bits The corresponding decoded value.

[0131] Synchronize the restored value with the sender Calculate the Hamming distance and get the consistency score . Extract path metrics at the same time , normalized to , fusion score:

[0132] ;

[0133] The final synchronization starting point is the highest scoring position If decoding fails after the synchronization starting point, the next best candidate point will be tried in order of confidence.

[0134] At the synchronization start point The full SCL list decoding is performed at the node with a path width of L=8 and a CRC check. If the check passes, the information bit is output; if it fails, the candidate point is switched.

[0135] On the basis of the same channel conditions and consistent code length, compared with the traditional frame synchronization scheme, the present invention shows: the synchronization accuracy is significantly improved; the average synchronization judgment time is significantly advanced; the system does not occupy additional bit resources, and the channel utilization rate is effectively improved.

[0136] Finally, it should be noted that the above only describes the preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art will appreciate that the technical solutions described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalent ones. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A self-synchronous decoding method based on polar codes, characterized in that, The application relates to a method for constructing a polar code encoding structure, and a method for decoding a polar code. In the process of constructing a polar code encoding structure at a sending end, a preset synchronization identification sequence is embedded in selected frozen bit positions, the frozen bits are set as predefined nonzero values on the premise of keeping a non-information bearing role, and the frozen bits are used for synchronization auxiliary identification at a decoding end; In the process of receiving a polar code encoding bit stream at a receiving end, based on the symbol stability change trend of the frozen bits in path expansion and the mutation characteristics of path metric values, a plurality of synchronization candidate starting positions are identified in real time; Specifically, the symbol stability change trend comprises: At the receiving end, the decoding values of the frozen bits in the decoding path are continuously monitored, and in each sliding window, the symbol consistency change of the frozen bit positions in the path list in the window is counted; If the symbol consistency of the frozen bits in a certain sliding window significantly decreases, that is, the decoding values of most frozen bits in the candidate paths are significantly inconsistent, and the symbol consistency decrease amplitude continuously exceeds a set amplitude threshold, the current window starting position is determined as a synchronization candidate starting point; Specifically, the mutation behavior of the path metric values comprises: In the decoding process, the path metric change trend of each candidate path is continuously monitored, when the path average metric value in the current time corresponding path set changes by more than a set change threshold limit compared with the last time, or the number of effective paths reserved in the path expansion process decreases by more than a set proportion in a single bit decision, the position corresponding to the decoding time is determined as a synchronization candidate starting position; A restricted path decoding operation is performed on each of the synchronization candidate starting positions, the frozen bit restoration values are extracted, and the frozen bit restoration values are compared with a preset synchronization identification sequence to obtain a consistency score; The path metric values and the consistency scores are weighted and fused to construct a synchronization confidence function, the position with the highest score is selected as a synchronization starting point, a complete polar code decoding process is performed starting from the synchronization starting point, and a restored information bit sequence is output.

2. The self-synchronous decoding method based on polar codes according to claim 1, characterized in that, The synchronization identification sequence is composed of a binary sequence with a fixed length, the sequence is generated by a linear feedback shift register, is embedded in selected frozen bit positions, and is distributed in a bit interval mode to avoid significant interference on the overall structure performance of the polar code.

3. The method of claim 1, wherein, The position selection of the frozen bits comprises the following steps: According to the code length parameter of the polar code, the reliability of each bit channel is estimated by using the Bartlett parameter method, and a bit channel reliability sorting list is generated; According to a set code rate, bit positions with a Bartlett parameter value lower than a first preset threshold are selected as information bits from the sorting list, and the remaining positions are selected as a frozen bit set; In the frozen bit set, a plurality of frozen bits with a Bartlett parameter value higher than a second preset threshold are selected as embedding positions of the synchronization identification sequence, and are uniformly distributed in the code word in a fixed interval mode; The bit values of the preset synchronization identification sequence are embedded into the embedding positions, so that the part of the frozen bits have a synchronization auxiliary function and do not bear effective information.

4. The self-synchronous decoding method based on polar codes according to claim 1, characterized in that, The consistency score acquisition step comprises: For each candidate starting position, a polar code partial path decoding operation is performed to obtain the restoration values of the frozen bits in the current path. The frozen bit restoration value is compared with the synchronization identification sequence embedded by the sending end bit by bit, and a Hamming distance is calculated as a consistency score; The smaller the Hamming distance is, the higher the matching degree is, and the more possible the corresponding candidate starting position is the real synchronization starting point.

5. The method of claim 1, wherein, The synchronization confidence function includes: Based on path metric information and frozen bit consistency score information, fusion is constructed; The path metric information is used to reflect the overall confidence degree of the decoding path, and a relative confidence ranking is formed among different candidate paths. The consistency score is used to measure the consistency degree of the frozen bit in the decoding path and the preset synchronization identification sequence, and the consistency score is obtained by performing a limited depth path decoding on each candidate synchronization starting position. The two types of information are fused according to a preset weighting strategy to form a comprehensive score index for synchronization starting point determination. The comprehensive score result is used to sort a plurality of candidate synchronization starting positions, and the position with the highest score is selected as the final synchronization starting point.

6. The method of claim 1, wherein, The complete polar code decoding process includes: The SCL method is used, and the correctness of a plurality of candidate decoding paths is judged in combination with the CRC. The effectiveness verification step of the synchronization starting point includes: after the polar code list decoding is completed at the selected synchronization starting position, if there is at least one path that passes the CRC check in the decoding path set, it is confirmed that the current synchronization starting point is an effective synchronization starting point, and the corresponding decoding result is output. If all paths do not pass the CRC check, the current synchronization starting point is abandoned, the next candidate synchronization position is selected in the order of the synchronization confidence score, and the complete decoding and verification process is re-executed. When all candidate synchronization positions do not pass the CRC check, it is determined that the current data block synchronization fails, and the next identifiable synchronization block is waited for.

Images