A polar code concatenated check decoding method without feedback channel

Abstract

The application discloses a polar code cascade check decoding method without feedback channel and relates to the technical field of communication and information security, and comprises the following steps: step one, according to the data type and the link estimated quality, the original data is subjected to data block segmentation, and the preliminary encoding parameters are determined, the CRC check code is generated for each data block, and is used as a part of the information bit to participate in the polarization coding; through the adaptive frozen bit selection, the enhanced CRC-aided SCL decoding and the confidence evaluation mechanism, the error correction performance is superior to that of the traditional RS code, Turbo code, LDPC code and the polarization code scheme with a fixed strategy under the condition of a short code length and a low signal-to-noise ratio, is suitable for a one-way network environment, does not need the ACK / NACK and retransmission mechanism, simplifies the network architecture, reduces the communication overhead and delay, and can dynamically adjust the transmission strategy according to the link state and the data characteristics, and can maintain the efficient and reliable transmission in various complex one-way network scenes.

Description

Technical Field

[0001] This invention relates to the field of communication and information security technology, specifically to a polar code concatenation verification and decoding method without feedback channels. Background Technology

[0002] One-way network environments, where data can only flow from the sender to the receiver without the ability to reverse acknowledgment or retransmission, are widely used in scenarios with special requirements for security or one-wayness, such as industrial control security gateways, satellite downlinks, broadcast television systems, military communications, and multicast distribution systems. In one-way networks, traditional error control mechanisms, such as CRC, checksums, and parity checks, can effectively detect errors, but because they cannot be retransmitted, the data is still unusable after an error is detected. Polar codes, as an emerging forward error correction code, have shown great potential due to their theoretical attainment of the Shannon limit, low encoding and decoding complexity, and application in the control channel of the 5G-NR standard. However, in the specific scenario of unidirectional networks, most existing polar code-based schemes adopt fixed frozen bit selection strategies, failing to fully consider the dynamic changes in unidirectional network link characteristics or the differences in reliability requirements of different types of data. This results in suboptimal performance in some scenarios. At the same time, existing technologies mainly rely on traditional continuous elimination or list decoding, lacking deep adaptation to the feedback-free characteristics of unidirectional networks. Although some studies have combined CRC with SCL, it is difficult to use CRC for more effective path selection assistance or error mitigation or data reconstruction in unidirectional network scenarios. Furthermore, it fails to dynamically adjust coding parameters according to link quality, data type, and real-time requirements, making it difficult to maintain efficient and reliable transmission in complex and ever-changing unidirectional network environments.

[0003] Therefore, in view of this, the present invention proposes a polar code concatenated verification decoding method without feedback channel to make up for and improve the shortcomings of the prior art. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a polar code concatenated verification and decoding method for a feedback-free channel that achieves high reliability in data transmission by leveraging the forward error correction capability of polar codes and innovative verification, decoding, and adaptive mechanisms. This solves the corresponding technical problems mentioned in the background section.

[0005] To achieve the above objectives, the technical solution adopted by this invention is: a polar code concatenated verification decoding method without feedback channel, comprising the following steps: Step 1: Based on the data type and estimated link quality, the raw data is divided into data blocks, and preliminary coding parameters are determined, including code length N and information bit length K. A CRC check code is generated for each data block and used as part of the information bits in polar coding to enhance the path decision capability during decoding. Step 2: Based on the preset link noise statistical characteristics, including bit error rate (BER) estimation and signal-to-noise ratio (SNR) range, dynamically adjust the frozen bit set F, and fill the coding vector U with the information bits containing CRC and the frozen bits. The frozen bits are known fixed values. This is achieved using the planned transformation matrix. Encode the vector U to generate a polar codeword X of length N; Step 3: Modulate the encoded bit sequence X and transmit it from the transmitter to the receiver through a one-way network. The receiver demodulates the received analog signal and calculates the log-likelihood ratio (LLR) for each bit. Step 4: Using list decoding as the basis, the list size can be dynamically adjusted according to the device capabilities and real-time requirements. During the SCL decoding process, CRC check information is used to filter and prune the generated candidate paths, and a confidence evaluation mechanism is introduced to comprehensively evaluate multiple candidate paths. Step 5: Perform CRC verification on the candidate data blocks of the decoded output, and reassemble the recovered data blocks in order to output a complete data stream.

[0006] As a preferred option, In step five, If a path passes the CRC check, select the path with the highest confidence level as the final data. If all candidate path CRC checks fail, an error mitigation mechanism is activated, selecting the path with the highest confidence and attempting limited error correction based on the redundancy information of the data block, or marking the data block as an error and reporting it to the upper layer. Although there are cases where one-way networks cannot report, it can still serve as an internal log or a basis for subsequent statistical analysis.

[0007] Preferably, the transmitting end includes an adaptive data blockization module, a CRC generation and embedding module, an adaptive polar code encoding module, and a modulation and transmission module; The adaptive data blockization module is used to input the original data stream, link quality indicator, and data type identifier. The link quality indicator is obtained from external measurements or historical data. Based on the data type and link quality, it dynamically determines the block size K and encoding code length N, and outputs the block-divided data bit stream. And n is a positive integer; The CRC generation and embedding module is used to input the segmented data bit stream, generate a CRC checksum for each data block, append the CRC checksum to the original data block to form a new information bit sequence, and output the original information bit sequence containing the CRC, with a length of [length missing]. ; The modulation and transmission module is used to input polar codeword X, modulate the binary bit stream X into a signal waveform suitable for unidirectional channel transmission, and transmit it through a unidirectional network interface.

[0008] Preferably, the adaptive polar code encoding module includes a frozen bit dynamic selection submodule, an information bit and frozen bit filling submodule, and a polarization transformation encoding submodule. The frozen bit dynamic selection submodule is used to input link noise characteristics, code length N, and information bit length. Based on channel polarization characteristics, sub-channel indices that are expected to have lower reliability under given link conditions are preferentially assigned to the frozen bits, using the following strategy: Based on different typical channel models and SNR ranges, multiple sets of frozen bit sets are pre-calculated and stored. The transmitting end selects the most suitable frozen bit table based on the actual perceived link quality. If a certain computational overhead is allowed, the reliability of each sub-channel can be dynamically calculated based on the real-time SNR estimate using methods such as Gaussian approximation, and the frozen bit can be selected accordingly. For high-priority control information, more or more critical sub-channels can be forced to be set as information bits, even if their theoretical reliability is slightly lower. Output the determined set of frozen bits F and its corresponding sub-channel index; The information bit and frozen bit filling submodule is used to take as input an information bit sequence containing CRC, a frozen bit set F, and a code length N, and fill the information bits into the non-frozen bit subchannel and the frozen bits into the frozen bit subchannel to output the encoding vector U to be encoded, with a code length of N, wherein the total number of information bits is N. ; The polarization transform coding submodule is used to input the coding vector U and the code length N, and to construct the polarization transform matrix recursively. Encode the encoding vector U to obtain the encoded codeword. , to output a polar codeword X of length N.

[0009] Preferably, the receiving end includes a signal receiving and demodulation module, an enhanced polar code decoding module, a CRC check and data recovery module, and a data reconstruction and output module; The signal receiving and demodulation module is used to input a one-way channel received signal, perform filtering, amplification, synchronization, and demodulation processing on the received signal, output hard decision bits or soft information, and output demodulated bit information or LLR sequence. The CRC check and data recovery module is used to input candidate data blocks, perform CRC check on the candidate data blocks, select the final data block based on the check result and confidence level evaluation, and if the selected data block is a complete sequence containing CRC, then the CRC is removed from it, the original data block is recovered, and the recovered original data block is output. The data reconstruction and output module is used to input the recovered data blocks, reconstruct the data blocks in their original order, and output a complete data stream. For data blocks marked as errors or unrecoverable, zero padding, interpolation, or error reporting are performed as needed.

[0010] Preferably, the enhanced polar code decoding module includes a CRC-assisted path pruning submodule and a confidence assessment and error mitigation submodule; The enhanced polar code decoding module uses the standard SCL decoding algorithm and maintains a path list of size L, where each path contains decoded bits and a partial path metric. The CRC-assisted path pruning submodule is used to obtain the codewords decoded from each path during the SCL decoding process and perform CRC verification on them as follows: If the validation fails, the path is pruned (unless the list is not full and there are no other paths available). If the verification passes, the path is preserved and can be given higher priority or used to guide the pruning of other paths; The confidence assessment and error mitigation submodule is used to calculate the overall confidence of each candidate path, comprehensively considering the path metric, the average or minimum absolute value of LLR in the unfrozen bits, and the "closeness" of the CRC check result, and performs the following optimal path selection and error mitigation attempts: When multiple paths pass CRC checks, the one with the highest confidence is selected; when all paths fail CRC checks, the path with the highest confidence is selected as the "most likely correct" output, and internal error counting and alarms are triggered. When there are paths with high confidence but still failing CRC, attempt limited error correction and repair using redundant information within the data block.

[0011] Compared with existing technologies, the beneficial effects of this invention are as follows: Through adaptive frozen bit selection, enhanced CRC-Aided SCL decoding, and confidence assessment mechanisms, the error correction performance is superior to traditional RS codes, Turbo codes, LDPC codes, and fixed-strategy polar code schemes under short code length and low signal-to-noise ratio conditions. It is fully applicable to unidirectional network environments, requires no ACK / NACK or retransmission mechanisms, simplifies the network architecture, reduces communication overhead and latency, and can dynamically adjust the transmission strategy according to link status and data characteristics. It can maintain efficient and reliable transmission in various complex unidirectional network scenarios. At the same time, the lightweight encoding and decoding structure and parameter configurability provided make it easy to deploy and apply in high-security, low-power scenarios such as embedded devices, satellite terminals, and industrial gateways. It organically combines "error detection" and "error correction" and is supplemented by intelligent error assessment and decision-making mechanisms, which greatly improves the availability and reliability of unidirectional data transmission. It can be widely used in various unidirectional network scenarios such as satellite downlink, broadcast television, industrial control gateways, military communications, IoT downlink control, and large-scale multicast distribution. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of the overall structure of a preferred embodiment of the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] Embodiments of the present invention: Please refer to Figure 1 As shown, a polar code concatenated verification decoding method without feedback channel includes the following steps: Step 1: Based on the data type and estimated link quality, the raw data is divided into data blocks, and preliminary coding parameters are determined, including code length N and information bit length K. A CRC check code is generated for each data block and used as part of the information bits in polar coding to enhance the path decision capability during decoding. Step 2: Based on the preset link noise statistical characteristics, including bit error rate (BER) estimation and signal-to-noise ratio (SNR) range, dynamically adjust the frozen bit set F, and fill the coding vector U with the information bits containing CRC and the frozen bits. The frozen bits are known fixed values. This is achieved using the planned transformation matrix. Encode the vector U to generate a polar codeword X of length N; Step 3: Modulate the encoded bit sequence X and transmit it from the transmitter to the receiver through a one-way network. The receiver demodulates the received analog signal and calculates the log-likelihood ratio (LLR) for each bit. Step 4: Using list decoding as the basis, the list size can be dynamically adjusted according to the device capabilities and real-time requirements. During the SCL decoding process, CRC check information is used to filter and prune the generated candidate paths, and a confidence evaluation mechanism is introduced to comprehensively evaluate multiple candidate paths. Step 5: Perform CRC verification on the candidate data blocks of the decoded output, and reassemble the recovered data blocks in order to output a complete data stream.

[0015] In step five, If a path passes the CRC check, select the path with the highest confidence level as the final data. If all candidate path CRC checks fail, an error mitigation mechanism is activated, selecting the path with the highest confidence and attempting limited error correction based on the redundancy information of the data block, or marking the data block as an error and reporting it to the upper layer. Although there are cases where one-way networks cannot report, it can still serve as an internal log or a basis for subsequent statistical analysis.

[0016] The transmitting end includes an adaptive data blockization module, a CRC generation and embedding module, an adaptive polar code encoding module, and a modulation and transmission module; The adaptive data blockization module takes the raw data stream, link quality indicator, and data type identifier as input. The link quality indicator is obtained from external measurements or historical data. Based on the data type and link quality, it dynamically determines the block size K and encoding code length N, and outputs the block-divided data bit stream. And n is a positive integer; The CRC generation and embedding module takes the segmented data bitstream as input, generates a CRC checksum for each data block, appends the CRC to the original data block to form a new information bit sequence, and outputs the original information bit sequence containing the CRC, with a length of [length missing]. ; The modulation and transmission module is used to input polar codeword X, modulate the binary bit stream X into a signal waveform suitable for unidirectional channel transmission, and transmit it through the unidirectional network interface.

[0017] The adaptive polar code encoding module includes a frozen bit dynamic selection submodule, an information bit and frozen bit filling submodule, and a polarization transformation encoding submodule; The frozen bit dynamic selection submodule is used to input link noise characteristics, code length N, and information bit length. Based on channel polarization characteristics, sub-channel indices that are expected to have lower reliability under given link conditions are preferentially assigned to the frozen bits, using the following strategy: Based on different typical channel models and SNR ranges, multiple sets of frozen bit sets are pre-calculated and stored. The transmitting end selects the most suitable frozen bit table based on the actual perceived link quality. If a certain computational overhead is allowed, the reliability of each sub-channel can be dynamically calculated based on the real-time SNR estimate using methods such as Gaussian approximation, and the frozen bit can be selected accordingly. For high-priority control information, more or more critical sub-channels can be forced to be set as information bits, even if their theoretical reliability is slightly lower. Output the determined set of frozen bits F and its corresponding sub-channel index; The information bit and frozen bit filling submodule takes as input an information bit sequence containing CRC, a frozen bit set F, and a code length N. It fills the information bits into the non-frozen bit subchannel and the frozen bits into the frozen bit subchannel to output the encoding vector U to be encoded, with a code length of N. The total number of information bits is... ; The polarization transform coding submodule is used to take the coding vector U and code length N as inputs and constructs the polarization transform matrix recursively. Encode the encoding vector U to obtain the encoded codeword. , to output a polar codeword X of length N.

[0018] The receiver includes a signal receiving and demodulation module, an enhanced polar code decoding module, a CRC check and data recovery module, and a data reconstruction and output module; The signal receiving and demodulation module is used to input unidirectional channel received signals, perform filtering, amplification, synchronization and demodulation processing on the received signals, and output hard decision bits or soft information, and output demodulated bit information or LLR sequence. The CRC check and data recovery module is used to input candidate data blocks, perform CRC check on the candidate data blocks, select the final data block based on the check result and confidence level evaluation, and if the selected data block is a complete sequence containing CRC, then the CRC is removed from it, the original data block is recovered, and the recovered original data block is output. The data reconstruction and output module is used to input the recovered data blocks, reconstruct the data blocks in their original order, and output a complete data stream. For data blocks marked as errors or unrecoverable, zero padding, interpolation, or error reporting are performed as needed.

[0019] The enhanced polar code decoding module includes a CRC-assisted path pruning submodule and a confidence assessment and error mitigation submodule; The enhanced polar code decoding module uses the standard SCL decoding algorithm and maintains a path list of size L, where each path contains decoded bits and a partial path metric. The CRC-assisted path pruning submodule is used to obtain the codewords decoded from each path during the SCL decoding process and perform CRC checks on them as follows: If the validation fails, the path is pruned (unless the list is not full and there are no other paths available). If the verification passes, the path is preserved and can be given higher priority or used to guide the pruning of other paths; The confidence assessment and error mitigation submodule is used to calculate the overall confidence of each candidate path, comprehensively considering the path metric, the average or minimum absolute value of LLR in the unfrozen bits, and the "closeness" of the CRC check results, and performs the following optimal path selection and error mitigation attempts: When multiple paths pass CRC checks, the one with the highest confidence is selected; when all paths fail CRC checks, the path with the highest confidence is selected as the "most likely correct" output, and internal error counting and alarms are triggered. When there are paths with high confidence but still failing CRC, attempt limited error correction and repair using redundant information within the data block.

[0020] In this embodiment: The specific implementation method for dynamic selection of the freeze position is as follows: Predefined frozen bit table method: For a typical one-way channel model, under different SNR points, the optimal set of frozen bits for the corresponding code length N is pre-calculated through Monte Carlo simulation, i.e., the set with the worst reliability. Each sub-channel index is stored in the non-volatile memory of the transmitter. The transmitter estimates the SNR of the current link in some way (such as pilot signals or historical bit error statistics), and then selects the set of frozen bits corresponding to the closest SNR from the predefined table. Dynamic calculation method based on Gaussian approximation: Assuming that the noise of each sub-channel is independent and identically distributed, the Bavarian parameters of each sub-channel are estimated using the Gaussian approximation method. Based on the current SNR estimate, calculate the SNR of all sub-channels. ,Will Sort the values ​​in ascending order. A smaller value indicates a more reliable channel; before selection... The least reliable subchannel is used as the freeze bit; Data type-driven adjustment: For data blocks marked as "emergency control instructions", based on the above frozen bit selection results, M of the less important (but theoretically more reliable) frozen bits can be forcibly restored to information bits in exchange for higher transmission efficiency, provided that the system can tolerate the slight reliability loss caused by this.

[0021] The specific implementation method of the polar code encoding module is as follows: Polarization transformation matrix Construction: Using a recursive formula , ,in, Indicates the Kronecker product; Low-complexity implementation: Polar coding is implemented using a butterfly operation structure, avoiding large-scale matrix multiplication operations; Bit reversal sorting: Reversing the bits of a bit sequence before or after encoding to meet specific implementation standards or optimization requirements.

[0022] The specific implementation of the enhanced polar code decoding module is as follows: SCL Decoding Basics: The standard SCL algorithm is used to maintain a path list of size L, where each path contains the decoded bits (or a portion of the bits) and the path metric. CRC-assisted path pruning: Assuming the CRC length is c, when a path is decoded... After each bit of information containing the CRC, it is organized into a data block, the CRC is calculated and compared with the additional CRC. If they do not match, the path is deleted. Example of confidence level assessment metrics: The ratio of a path metric to the minimum path metric in the list; the closer the ratio is to 1, the higher the confidence level. For the decoded information bits, calculate the sum or average of their absolute LLR values. The larger the value, the higher the confidence level. For CRC check, the Hamming distance between the received CRC bits and the calculated CRC bits can be calculated. The smaller the distance, the greater the probability that the CRC is correct. Adaptive adjustment of list size L: Initially set to 4 or 8. If the CRC check failure rate is high in decoding several consecutive data blocks, L is dynamically increased to 16 or 32; conversely, if the success rate is high, L can be appropriately reduced to reduce power consumption. Example of an error mitigation mechanism: When all L paths fail CRC checks, select the path with the smallest path metric for output, and record the block number and error flag for subsequent analysis.

[0023] The specific implementation method for CRC check path selection and data recovery is as follows: Path selection logic: function select_best_path(candidate_paths): passed_paths = [] for each path in candidate_paths: if path.crc_check_passed(): passed_paths.append(path) if len(passed_paths) > 0: return path_with_highest_confidence(passed_paths) / / For example, the path with the lowest confidence. else: return path_with_highest_confidence(candidate_paths) / / The path with the highest confidence among all paths Data Recovery: Extract the original data from the selected path The K bits contain the CRC information bits, and the CRC part is stripped from them to obtain the K-bit original data block.

[0024] Furthermore, in this embodiment, the system parameter configuration example is as follows: Control information transmission: Data type identifier: control command; Code length N: 64 bits; Information bits 32 bits (including 24 bits of data + 8 bits of CRC-8); Freeze position 32 bits; SCL list size L: 8; Modulation method: BPSK; Service data transmission (good channel): Data type identifier: Normal business; Code length N: 256 bits; Information bits 192 bits (including 180 data bits + 12 CRC-12 bits); Freeze position 64 bits; SCL list size L: 4; Modulation method: QPSK; Business data transmission (poor channel): Data type identifier: Normal business; Code length N: 512 bits; Information bits 256 bits (including 240 data bits + 16-bit CRC-16); Freeze position 256 bits; SCL list size L: 16; Modulation method: BPSK.

[0025] This invention can be widely applied to various one-way network environments, including but not limited to: Industrial control systems: such as safety gates in industries like power and petrochemicals, used to send control commands and configuration parameters to lower-level machines or PLCs; Satellite communication: Satellite downlink broadcasts multimedia information such as data, images, and video to ground stations or mobile terminals; Broadcast television system: content transmission for digital television and audio broadcasting; Military communications: the one-way distribution of tactical command instructions and intelligence information; Internet of Things (IoT): Base stations send configuration information and control commands to a massive number of sensor nodes; Massive multicast distribution systems: such as the distribution of software updates, news pushes, and electronic newspapers; Emergency communication system: When a disaster occurs, the rescue command center sends out early warning information and rescue instructions unidirectionally to the affected area.

[0026] By employing a dynamic optimization selection mechanism for frozen bits based on link awareness and data type drive, this approach breaks through the rigid selection mode of frozen bits in existing polar code schemes. It not only considers the inherent reliability ranking of sub-channels but also introduces real-time or near-real-time link noise characteristics (such as SNR and BER) and the importance / timeliness of data types as key criteria for frozen bit selection. For example, during periods of high noise, more sub-channels with slightly lower reliability but theoretically still potential will be set as frozen bits to protect core information. For emergency control commands, the number of information bits in some secondary sub-channels may be sacrificed to improve transmission efficiency while ensuring a certain level of reliability. This enables "on-demand polarization" of polar codes in dynamic unidirectional network environments, maximizing channel polarization gain and significantly improving average error correction performance in different scenarios.

[0027] An enhanced SCL decoding and error detection method that integrates CRC check and confidence assessment is introduced on the basis of CRC-AidedSCL. It further introduces a multi-dimensional confidence assessment system. In addition to the CRC check result, it also considers information such as path metric and LLR confidence to perform more refined sorting and selection of decoding paths. Even when all candidate paths fail CRC check, it can still output the "best guess" result based on confidence and provide error indication. This solves the problem that traditional CRC-Aided SCL may have "no path to choose" under extremely poor channel conditions. As a result, it significantly improves the robustness of decoding and the availability of output data under one-way network without feedback conditions, and effectively reduces the false negative rate and false positive rate.

[0028] By proposing an adaptive frame structure and dynamic parameter configuration mechanism for unidirectional networks, a complete adaptive parameter adjustment strategy is proposed. The system can dynamically adjust the following parameters based on link quality monitoring, data type identification, and system resource constraints (such as buffer size and processing latency requirements): the coding block size K and code length N (affecting the code rate R=K / N), the length and polynomial selection of the CRC checksum, the list size L of the SCL decoding, and the aggressiveness of the freeze bit selection strategy. This enables the system to achieve the best balance between efficiency (high code rate) and reliability (low code rate) and flexibly adapt to the application requirements of different types of unidirectional networks.

[0029] By proposing a lightweight, configurable polar code codec architecture and its low-power implementation in unidirectional network nodes, this paper proposes a modular and customizable polar code codec implementation architecture for resource-constrained devices such as embedded systems, security gateways, and satellite terminals. For example, the encoding end can adopt a butterfly arithmetic unit array based on shift registers and XOR gates; the decoding end SCL can adopt pipelined design and storage optimization techniques. At the same time, a variety of optional simplified algorithms (such as LLR update based on approximate calculation and simplified confidence metric function) are provided, allowing users to customize the functionality according to the device's computing power and power budget. This reduces the deployment threshold of polar code technology in unidirectional network edge devices and embedded systems, and promotes its widespread application.

[0030] By establishing an integrated data integrity assurance closed loop of "detection-error correction-evaluation-decision" in a feedback-free network, this invention does not simply superimpose polar codes and CRC. Instead, it constructs a complete closed-loop system from data preprocessing (including CRC), adaptive encoding (including dynamic freezing bits), enhanced decoding (including CRC assistance and confidence assessment) to final data recovery and error handling. This closed loop fully considers the feedback-free characteristics of unidirectional networks and maximizes data integrity through internal adaptive mechanisms and error mitigation strategies. It fundamentally changes the passive situation of "detecting errors and discarding data" in unidirectional network data transmission and achieves a leap from "active error correction and intelligent fault tolerance".

[0031] This application creatively proposes a dynamic optimization selection mechanism for frozen bits based on link awareness and data type driving, an enhanced SCL decoding method that integrates CRC check and confidence assessment, an adaptive frame structure and parameter dynamic configuration mechanism for unidirectional networks, and a lightweight encoding and decoding implementation scheme. It constructs a logically closed-loop, highly adaptive data integrity guarantee system. This method effectively overcomes the shortcomings of traditional unidirectional network data transmission, such as the inability to correct errors after detection and the insufficient performance or excessive complexity of traditional error correction codes. It significantly improves the reliability of data transmission under short code length and no feedback conditions.

[0032] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.

[0033] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. In the two embodiments provided in this application, it should be understood that the disclosed apparatus and system can be implemented in other ways; for example, the apparatus embodiments described above are merely illustrative, and the division of modules is merely a logical functional division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed; furthermore, the coupling or direct coupling or communication connection between the shown or discussed mutuals can be through some interfaces, and the indirect coupling or communication connection between the apparatus or modules can be electrical, mechanical or other forms. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A polar code concatenated verification decoding method without feedback channel, characterized in that, Includes the following steps: Step 1: Based on the data type and estimated link quality, the raw data is divided into data blocks, and preliminary coding parameters are determined, including code length N and information bit length K. A CRC check code is generated for each data block and participates in polar coding as part of the information bits. Step 2: Based on the preset link noise statistical characteristics, dynamically adjust the frozen bit set F, and fill the information bits and frozen bits into the encoding vector U, using the planned transformation matrix. Encode the vector U to generate a polar codeword X of length N; Step 3: Modulate the encoded bit sequence X and transmit it from the transmitter to the receiver through a one-way network. The receiver demodulates the received analog signal and calculates the log-likelihood ratio of each bit. Step 4: Using list decoding as a basis, CRC check information is used to filter and prune the generated candidate paths, and a confidence evaluation mechanism is introduced to comprehensively evaluate multiple candidate paths. Step 5: Perform CRC verification on the candidate data blocks of the decoded output, and reassemble the recovered data blocks in order to output a complete data stream.

2. The polar code concatenated verification decoding method without feedback channel according to claim 1, characterized in that, In step five, If a path passes the CRC check, select the path with the highest confidence level as the final data. If all candidate path CRC checks fail, an error mitigation mechanism is activated, selecting the path with the highest confidence and attempting limited error correction based on the redundancy information of the data block.

3. The polar code concatenated verification decoding method without feedback channel according to claim 1, characterized in that, The transmitting end includes an adaptive data blockization module, a CRC generation and embedding module, an adaptive polar code encoding module, and a modulation and transmission module; The adaptive data blockization module is used to input the original data stream, link quality indicator, and data type identifier, dynamically determine the block size K and code length N according to the data type and link quality, and output the block-divided data bit stream. The CRC generation and embedding module is used to input the data bit stream after it is divided into blocks, generate a CRC check code for each data block, append the CRC code to the original data block to form a new information bit sequence, and output the original information bit sequence containing the CRC. The modulation and transmission module is used to input polar codeword X, modulate the binary bit stream X into a signal waveform suitable for unidirectional channel transmission, and transmit it through a unidirectional network interface.

4. The polar code concatenated verification decoding method without feedback channel according to claim 3, characterized in that, The adaptive polar code encoding module includes a frozen bit dynamic selection submodule, an information bit and frozen bit filling submodule, and a polarization transformation encoding submodule. The frozen bit dynamic selection submodule is used to input link noise characteristics, code length N, and information bit length. Based on the channel polarization characteristics, the sub-channel indices that are expected to have lower reliability under given link conditions are preferentially assigned to the frozen bits, so as to output a determined set of frozen bits F and its corresponding sub-channel indices. The information bit and frozen bit filling submodule is used to take as input an information bit sequence containing CRC, a frozen bit set F, and a code length N, fill the information bits into the non-frozen bit subchannel, and fill the frozen bits into the frozen bit subchannel, so as to output the encoding vector U to be encoded, wherein the total number of information bits is... ; The polarization transform coding submodule is used to input the coding vector U and the code length N, and to construct the polarization transform matrix recursively. Encode the encoding vector U to obtain the encoded codeword. , to output a polar codeword X of length N.

5. The polar code concatenated verification decoding method without feedback channel according to claim 1, characterized in that, The receiving end includes a signal receiving and demodulation module, an enhanced polar code decoding module, a CRC check and data recovery module, and a data reconstruction and output module; The signal receiving and demodulation module is used to input a one-way channel received signal, perform filtering, amplification, synchronization, and demodulation processing on the received signal, output hard decision bit information, and output demodulated bit information. The CRC check and data recovery module is used to input candidate data blocks, perform CRC check on the candidate data blocks, select the final data block based on the check result and confidence level evaluation, and if the selected data block is a complete sequence containing CRC, then the CRC is removed from it, the original data block is recovered, and the recovered original data block is output. The data reconstruction and output module is used to input the recovered data blocks, reconstruct the data blocks in their original order, and output a complete data stream.

6. The polar code concatenated verification decoding method without feedback channel according to claim 5, characterized in that, The enhanced polar code decoding module includes a CRC-assisted path pruning submodule and a confidence assessment and error mitigation submodule; The enhanced polar code decoding module uses the SCL decoding algorithm and maintains a path list of size L, where each path contains decoded bits and a partial path metric. The CRC-assisted path pruning submodule is used to obtain the codewords decoded from each path and perform CRC verification on them as follows: If the verification fails, the path is pruned. If the verification passes, the path will be preserved. The confidence assessment and error mitigation submodule is used to calculate the overall confidence of each candidate path and to attempt optimal path selection and error mitigation as follows: When multiple paths pass CRC checks, the one with the highest confidence is selected; when all paths fail CRC checks, the path with the highest confidence is selected as the "most likely correct" output, and internal error counting and alarms are triggered. When there are paths with high confidence but still failing CRC, attempt limited error correction and repair using redundant information within the data block.

Images