Parallel stepped convolutional code space coupling coding and decoding method and device

Abstract

The invention discloses a parallel stepped convolutional code space coupling coding and decoding (PSCBCC) method and device, belongs to the technical field of channel coding, and is suitable for a large-bandwidth communication scene. According to the method, a frame of data is divided into multiple groups of sub-data, parallel convolutional coding of three paths of input of current sub-data, interleaved previous sub-data and an interleaved previous check sequence is realized by utilizing a parallel coding unit, a current check sequence is generated, and the sub-data and the check sequence are respectively reserved to different subsequent moments to form asymmetric coupling; and during decoding, a parallel decoding unit is matched with a coding structure, and a plurality of groups of log-likelihood ratios are input to realize iterative updating. The throughput rate is improved through parallel processing, the bit error rate performance is enhanced through asymmetric coupling, multi-code-rate adaptation is supported, and the performance is superior to that of existing BCC and 4G / 5G standard codes.

Description

A Parallel Ladder Convolutional Code Spatial Coupled Encoding and Decoding Method and Apparatus Technical Field

[0001] This invention relates to the field of channel coding technology, and discloses a parallel staircase convolutional code spatial coupling encoding and decoding method and apparatus. Background Technology

[0002] In high-bandwidth communication scenarios such as 6G wireless transmission and high-speed fiber optic communication, a single data frame typically contains tens of thousands of bits, requiring channel coding to reduce the transmission error rate and retransmission overhead. Spatial coupling coding technology has become the mainstream solution due to its performance approaching the Shannon limit, with multi-bit input systematic convolutional code concatenation spatial coupling coding (BCC) being a typical example.

[0003] Figure 1 shows a schematic diagram of the existing BCC encoder structure. , The term "BCC" stands for interleaver, and "ENC" stands for convolutional code encoder. A BCC is constructed by cascading multi-bit input system convolutional code encoders, where the encoder receives two or more information bit sequences as input. It processes a frame of data... Divided into Subgroup data The verification sequence output by the encoder in the previous moment will be... The encoder at each time step is used as input for re-encoding. Here, The coupling memory length, known as SCPCC, .

[0004] The core of the existing BCC scheme is to divide the data into multiple sets of sub-data. The verification sequence of the previous time step is reused by the encoder of the next m time steps (symmetric coupling). However, it has significant limitations: (1) Coupling symmetry leads to redundancy conflicts: the sub-data and the verification sequence share the same coupling memory length, and the information is repeatedly encoded at the same time. Redundancy overlap reduces the coding gain; (2) The parallelism is fixed and the throughput is limited: the encoding process is serial, and the encoding time of a single set of data is linearly related to the length, which is difficult to adapt to the high-speed requirements of large bandwidth scenarios; (3) The bit error rate performance is poor under low signal-to-noise ratio.

[0005] Therefore, there is an urgent need for a codec scheme with parallel processing capabilities, asymmetric coupling structure, and support for multiple code rates to improve the transmission reliability and efficiency in high-bandwidth scenarios. Summary of the Invention

[0006] The purpose of this invention is to provide a parallel staircase convolutional code spatial coupling encoding and decoding method and apparatus. By introducing parallel encoding / decoding units, the throughput is improved, and an asymmetric coupling structure is designed to enhance bit error rate performance. At the same time, it supports multi-code rate adaptation and meets the high reliability and efficiency requirements of high bandwidth communication scenarios.

[0007] To achieve the above objectives, the present invention provides the following technical solution: According to one aspect of the present invention, a parallel staircase convolutional code spatial coupling coding method is provided, comprising: dividing a frame of data into multiple groups of sub-data sequences; for each group of sub-data, performing encoding by a parallel coding unit: using the group of sub-data as a first input, using previously specific sub-data after interleaving as a second input, using previously specific parity sequence after interleaving as a third input, and generating a current parity sequence; wherein, the parallel coding unit divides the input into S sub-sequences through a demultiplexer, encodes them with S parallel convolutional codes, and then merges them into the current parity sequence by a multiplexer, where S is the number of parallel sequences and S≥1; outputting a codeword containing the group of sub-data and the current parity sequence; retaining the group of sub-data until a subsequent first specific time as the input of the parallel coding unit, and retaining the current parity sequence until a subsequent second specific time as the input of the parallel coding unit, wherein the first and second specific times are different; if the previously specific sub-data or parity sequence does not exist, it is replaced with an initial value.

[0008] According to one embodiment of the present invention, the number of the multiple sets of sub-data sequences is twice the number of the original data frame groups; the length of each set of sub-data is half the length of a single set of data in the original data frame, the length of the current verification sequence is the same as the length of a single set of data in the original data frame, and the total system code rate is 1 / 3; the value of the parallel number S satisfies that the length of the sub-sequence after the sub-data is divided is greater than or equal to 50 bits.

[0009] According to one embodiment of the present invention, the first specific time is the current time plus m-1, and the second specific time is the current time plus m, where m is an integer greater than 2; the convolutional code encoder in the parallel coding unit is a multi-bit input system convolutional code encoder with a code rate of 2 / 3.

[0010] According to one embodiment of the present invention, the parallel coding unit employs a multi-bit input system convolutional code encoder with a code rate of k / n, where n is the number of output bits, k is the number of input bits, and nk > 1 and 2k - n ≥ 1, wherein: the length of each group of sub-data is K, and the first and second input paths are merged into one. The path sequence; the third input is The path sequence; the length of the current verification sequence is The total system bitrate is .

[0011] On the other hand, the present invention also provides a parallel staircase convolutional code spatial coupling coding device, comprising: a data segmentation module for segmenting a frame of data into multiple sets of sub-data sequences; a parallel coding unit comprising a demultiplexer, an S-path convolutional code encoder, and a multiplexer, wherein the demultiplexer segments the current sub-data, the previous specific sub-data, and the previous specific parity sequence into S-path sub-sequences, the convolutional code encoder encodes the sub-sequences in parallel, and the multiplexer merges the encoding results to generate the current parity sequence; an output module for outputting codewords containing the current sub-data and the current parity sequence; a storage module for retaining the current sub-data until a subsequent first specific time and retaining the current parity sequence until a subsequent second specific time, wherein the first and second specific times are different; and providing an initial value when the previous specific sub-data or parity sequence does not exist.

[0012] On the other hand, the present invention also provides a parallel staircase convolutional code spatial coupling decoding method, comprising the following steps: performing iterative decoding on the codewords; for each group of sub-data corresponding to the codewords, inputting the channel log-likelihood ratio and extrinsic information of the group of sub-data, the previous specific sub-data, and the channel log-likelihood ratio and extrinsic information of the group of parity sequences and the previous specific parity sequences through a parallel decoding unit; wherein, the parallel decoding unit divides the input log-likelihood ratio into S-way sub-sequences through a demultiplexer, and after decoding with S-way parallel convolutional codes, the multiplexer merges them into updated extrinsic information; outputting the updated extrinsic information and the predicted value of the corresponding sub-data; performing decoding according to a preset iteration order, and outputting the final decoding result when all codewords are successfully decoded or the maximum number of iterations is reached.

[0013] According to one embodiment of the present invention, the input log-likelihood ratio includes 8 groups, namely: channel log-likelihood ratio of the current sub-data, channel log-likelihood ratio of the previous specific sub-data, extrinsic information of the current sub-data, extrinsic information of the previous specific sub-data, channel log-likelihood ratio of the current check sequence, channel log-likelihood ratio of the previous specific check sequence, extrinsic information of the current check sequence, and extrinsic information of the previous specific check sequence; wherein, the initial value of the extrinsic information is 0.

[0014] According to one embodiment of the present invention, the preset iteration order includes at least one of the following: sequential decoding from front to back, sequential decoding from back to front, parallel decoding, and sliding window decoding.

[0015] According to one embodiment of the present invention, the convolutional code decoder in the parallel decoding unit adopts a soft-input soft-output algorithm, including but not limited to the BCJR algorithm, the MAP algorithm, or the Log-MAP algorithm.

[0016] On the other hand, the present invention also provides a parallel ladder convolutional code spatial coupling decoding device, comprising: an iteration control module for controlling the iterative decoding process of codewords and setting a preset iteration order; an information input module for inputting the channel log-likelihood ratio and extrinsic information of sub-data and check sequences; a parallel decoding unit comprising a demultiplexer, an S-way convolutional code decoder, and a multiplexer, wherein the demultiplexer divides the input log-likelihood ratio into S-way sub-sequences, the decoder decodes the sub-sequences in parallel, and the multiplexer merges the decoding results to generate updated extrinsic information and sub-data prediction values; and a result output module for outputting the final decoding result when all codewords are successfully decoded or the maximum number of iterations is reached. Compared with the prior art, the beneficial effects of the present invention are: 1. 1. Increased throughput: Through parallel encoding / decoding units (parallelism S is adjustable), the processing time for a single data set is reduced to 1 / S. For example, when S=4, the processing time is 1 / 4 of that of serial processing, significantly improving transmission efficiency in high-bandwidth scenarios. 2. Superior bit error rate performance: Adopting an asymmetric coupling structure, different m1 and m2 avoid redundancy conflicts. At the same code rate, the bit error rate is reduced by more than one order of magnitude compared to traditional BCC, and performance is improved by more than 0.1dB at low signal-to-noise ratios. 3. Multi-code rate adaptation: Supports k / n code rates, allowing flexible application to diverse scenarios in 4G / 5G evolution and 6G. 4. Strong compatibility: Performance is superior to LTE-turbo and NR-LDPC standard codes, and the parallelism can be dynamically adjusted according to hardware resources (such as FPGA parallel processing capabilities), enabling flexible engineering implementation. Attached Figure Description

[0017] The accompanying drawings, which are provided to further illustrate the invention and constitute a part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 is a schematic diagram of a conventional BCC codec.

[0018] Figure 2 is a flowchart of the parallel staircase convolutional code spatial coupling coding method.

[0019] Figure 3 is a schematic diagram of a parallel staircase convolutional code spatially coupled coding device.

[0020] Figure 4 is a flowchart of the parallel staircase convolutional code spatial coupling decoding method.

[0021] Figure 5 is a schematic diagram of a parallel staircase convolutional code spatially coupled decoding device.

[0022] Figure 6 is a schematic diagram of the general structure of an encoder for spatially coupled parallel staircase convolutional coding.

[0023] Figure 7 is a schematic diagram of the encoder structure with m1=1 and m2=2.

[0024] Figure 8 is a schematic diagram of the general structure of a decoder for parallel staircase convolutional code spatial coupling decoding.

[0025] Figure 9 is a schematic diagram comparing the effects of the parallel staircase convolutional code spatial coupling encoding and decoding method of the present invention with those of the prior art. Detailed Implementation

[0026] To facilitate a clear description of the technical solutions in the embodiments of the present invention, the terms "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect. For example, the first threshold and the second threshold are merely used to distinguish different thresholds and do not limit their order. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" are not necessarily different.

[0027] It should be noted that in this invention, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0028] In this invention, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one" or similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, "at least one of a, b, or c" can represent: a, b, c, a combination of a and b, a combination of a and c, a combination of b and c, or a, b, and c, where a, b, and c can be single or multiple.

[0029] The Parallel Stepped Convolutional Code Spatial Coupled Encoding and Decoding (PSCBCC) method and apparatus of the present invention are suitable for high-bandwidth communication scenarios. This method divides a frame of data into multiple groups of sub-data. Parallel coding units are used to perform parallel convolutional coding of the current sub-data, interleaved previous sub-data, and interleaved previous parity sequence as three inputs, generating the current parity sequence. The sub-data and parity sequence are retained at different subsequent time points to form asymmetric coupling. During decoding, a parallel decoding unit matches the coding structure, and multiple sets of log-likelihood ratios are input for iterative updates. This invention improves throughput through parallel processing, enhances bit error rate performance through asymmetric coupling, and supports multi-code-rate adaptation, outperforming existing BCC and 4G / 5G standard codes.

[0030] Figure 2 shows a flowchart of the parallel staircase convolutional code spatial coupling coding method, including the following steps: Step 201, data segmentation: Divide a frame of data into 2L groups of sub-data sequences. The original data frame has L groups, and the length of each sub-data group is set according to the code rate; Step 202, parallel coding: For the t-th sub-data group, the parallel coding unit performs three-way input processing: First input: the current sub-data Second input: previously specific sub-data ,Delay At any given moment, after interleaving; the third input: a previously specified check sequence. ,Delay At any given time, after interleaving, the parallel coding unit divides the three inputs into S subsequences using a demultiplexer. These subsequences are then encoded in parallel by an S-channel convolutional encoder, and finally merged by a multiplexer to generate the current check sequence. Step 203, Asymmetric Preservation: ... Reserved for later At that moment, Reserved for later At that moment, All values ​​are greater than 0 and are used as input for subsequent encoding; if the previous data does not exist, it is replaced with an initial value (such as 0); Step 204, codeword output: the output includes and The code .

[0031] Figure 3 shows a schematic diagram of the encoding device. The encoding device includes a data segmentation module, a parallel encoding unit (splitter, S-channel encoder, multiplexer), an output module, and a storage module; wherein: the data segmentation module is used to segment a frame of data into multiple sets of sub-data sequences; the parallel encoding unit includes a splitter, an S-channel convolutional encoder, and a multiplexer, the splitter segmenting the current sub-data, the previous specific sub-data, and the previous specific check sequence into S-channel sub-sequences, the convolutional encoder encoding the sub-sequences in parallel, and the multiplexer merging the encoding results to generate the current check sequence; the output module is used to output codewords containing the current sub-data and the current check sequence; the storage module is used to retain the current sub-data until a subsequent first specific time point and the current check sequence until a subsequent second specific time point, and the first and second specific time points are different; when the previous specific sub-data or check sequence does not exist, an initial value is provided.

[0032] As shown in Figure 4, the parallel staircase convolutional code spatial coupling decoding method includes the following steps: Step 401, iterative initialization: for each group of codewords The parallel decoding unit receives 8 sets of log-likelihood ratios (LLRs), including: the current and previous sub-data, the channel LLR of the current and previous check sequences, and the extrinsic information LLR. The initial value of the extrinsic information is set to 0. Step 402, parallel decoding: the demultiplexer divides the input LLR into S-way sub-sequences, which are then processed in parallel by the S-way convolutional code decoder using a soft-input soft-output algorithm to output the updated extrinsic information LLR and the predicted sub-data values. Step 403, iterative control: the iteration is repeated in a preset order (such as bidirectional, parallel, or sliding window) until all codewords are successfully decoded or the maximum number of iterations is reached, and the final result is output.

[0033] Figure 5 shows a schematic diagram of the decoding device. The decoding device includes an iterative control module, an information input module, a parallel decoding unit (demultiplexer, S-channel decoder, and multiplexer), and a result output module, which respectively correspond to the functional implementation of the decoding method shown in Figure 4.

[0034] The iterative control module controls the iterative decoding process of the codewords and sets a preset iteration order; the information input module is used to input the channel log-likelihood ratio and extrinsic information of the sub-data and the check sequence; the parallel decoding unit includes a demultiplexer, an S-way convolutional code decoder, and a multiplexer. The demultiplexer divides the input log-likelihood ratio into S-way sub-sequences, the decoder decodes the sub-sequences in parallel, and the multiplexer merges the decoding results to generate updated extrinsic information and sub-data prediction values; the result output module outputs the final decoding result when all codewords are successfully decoded or the maximum number of iterations is reached.

[0035] Example 1: The encoding process of PSCBCC code is shown in Figure 6, which gives a schematic diagram of the general structure of the encoder. , The symbol represents an interleaver, ENC represents a convolutional code encoder, and the long bar represents a multiplexer (merging multiple data streams into one) or a demultiplexer (splitting one data stream into multiple streams). This represents a delay module that retains the current data for m units of time before outputting it (encoding a set of data). The time taken is 1 unit of time.

[0036] As shown in Figure 7, , A schematic diagram of the encoder structure. Taking a 2 / 3 bit-rate multi-bit input system convolutional code as an example, the data of one frame... Divided into Group data sequence For the first one Group data The encoding steps are as follows: First, the parallel encoding unit will... As the first input After being interleaved as the second input, After interleaving, the result is used as the third input, which is then encoded to obtain the check sequence. .

[0037] Secondly, the encoder ultimately outputs codewords. .

[0038] Then, retain the verification sequence. To the following The time step serves as the second input to the parallel coding unit, preserving the data sequence. To the following The time is used as the third input of the parallel coding unit.

[0039] Here, for any check sequence ,when hour, Initialize to 0. Similarly, for any data sequence... ,when hour, Initialize to 0. It is worth noting that in this coupled structure, the above encoding process must satisfy... ,and , All are greater than 0. Preferably, it can be made... and (must meet) ).

[0040] Assumption The length is , The length is , verification sequence The length is The total bitrate of the system is: .

[0041] In Figure 7, the encoding process of the parallel encoding unit is as follows: Input: , , Output: Key parameters: (Number of parallel lines).

[0042] Step 1: The demultiplexer will Divided into Subsequences of equal length .

[0043] Step 2: The demultiplexer will... Divided into Subsequences of equal length .

[0044] Step 3: The demultiplexer will... Divided into Subsequences of equal length .

[0045] Step 4: Proceed Parallel convolutional code encoding of the path data to obtain Path check sequence For the first The processing of the path data is as follows: Step 7a, the multiplexer will... and Merge into one data stream, denoted as .

[0046] Step 7b, the convolutional code encoder will As the first input, Encode the second input and output the checksum sequence. .

[0047] Step 5: The multiplexer will Merge them into one path, denoted as the check sequence. .

[0048] The number of parallel coding units in a convolutional code encoder (parallel number) (This value can be freely adjusted, with a minimum value of 1. Assume...) The length is Then each subsequence The length is Preferred, The length needs to be no less than 50 bits.

[0049] Example 2: Decoding Process of PSCBCC Code Any spatially coupled code requires iterative decoding, meaning that each codeword corresponding to a set of data needs to be decoded multiple times. Iterative decoding has several decoding order implementation schemes, the most direct of which is as follows: First, for the i-th iteration, decode each codeword in a forward-to-back order, then decode each codeword in a backward-to-forward order. Second, when all codewords are successfully decoded, or the maximum number of iterations is reached, the decoding process ends, and all data is output. Besides this scheme, there are other decoding order implementation schemes such as forward-to-back sequential decoding in each iteration, parallel decoding, and sliding window decoding. The BCC code designed in this invention is not limited to a specific codeword decoding order.

[0050] make and These represent the corresponding signals directly calculated from the received signals. and The channel log-likelihood ratio (channel LLR). Represents the corresponding output of the convolutional code decoder The external information log-likelihood ratio (external information LLR).

[0051] For the Group data The code The decoding process is as follows: First, for the codeword The decoding, the input of the parallel decoding unit , , , A total of 8 groups of LLRs.

[0052] Then, the parallel decoding unit outputs the updated extrinsic information LLR. , and for data Predicted value .

[0053] Figure 8 shows a schematic diagram of the general structure of a decoder. , Represents an interleaver. , The symbol represents a deinterleaver. DEC represents a convolutional code decoder, and the long bars represent multiplexers (merging multiple data streams into one) or demultiplexers (splitting one data stream into multiple streams).

[0054] enter: , , , Output: , . Key parameters: (Number of parallel lines).

[0055] The parallel decoding unit decoding process is as follows: Step 801, calculate , , , .

[0056] Step 802, the distributor will... Divide into subsequences The corresponding log-likelihood ratio subsequence .

[0057] Step 803, the splitter will interleave the... Divide into subsequences The corresponding log-likelihood ratio subsequence .

[0058] Step 804, the splitter will interleave the... Divide into subsequences The corresponding log-likelihood ratio subsequence .

[0059] Step 805, the distributor will Divide into subsequences The corresponding log-likelihood ratio subsequence .

[0060] Step 806, proceed Parallel convolutional code decoding of the path data. For the first... The processing of the path data is as follows: Step 806a, the multiplexer will... and The corresponding log-likelihood ratio subsequence and Merge into one route, denoted as .

[0061] Step 806b, input to the convolutional code decoder , , Output , , and for data Predicted value .

[0062] Step 806c, the demultiplexer will Segmentation Correspondence and Hard decision subsequence and .

[0063] Step 807, the multiplexer will Merge into one path to obtain data Predicted value .

[0064] Step 808, the multiplexer will The data is obtained by merging the two channels into one and then deinterleaving them. Predicted value .

[0065] Step 809, the multiplexer will Merge into one path, and get Step 810, the multiplexer will Merge into one path, and after deinterleaving, obtain Step 811, the multiplexer will Merge into one path, and after deinterleaving, obtain .

[0066] Step 812, the multiplexer will Merge into one path, and get .

[0067] The convolutional code decoder here can employ any convolutional code decoding algorithm that supports soft-in-soft-out (SISO), and is not limited to any specific algorithm. During the iterative decoding process, The extrinsic information LLR and the predicted value will be updated multiple times, and the decoder always retains only the latest extrinsic information LLR and the predicted value. (The initial value of the extrinsic information LLR is set to 0).

[0068] Example 3: Encoding Process of Extended PSCBCC Code The encoding structure of PSCBCC code can be extended to use a code rate of... It is constructed by concatenating multi-bit input system convolutional codes. Here , .make , . Data of one frame Divided into Group data sequence For the first one Group data The encoding steps are as follows: First, the parallel encoding unit will... As the first input After being interleaved as the second input, After interleaving, the result is used as the third input, which is then encoded to obtain the check sequence. .

[0069] Secondly, the encoder ultimately outputs codewords. .

[0070] Finally, retain the verification sequence. To the following The time step serves as the second input to the parallel coding unit, preserving the data sequence. To the following The time is used as the third input of the parallel coding unit.

[0071] enter: , , Output: Key parameters: The parallel encoding process of the parallel encoding unit is as follows: Step 901, the demultiplexer will... Divided into Subsequences of equal length Step 902, the distributor will... Divided into Subsequences of equal length Step 903, the distributor will... Divided into Subsequences of equal length Step 904, proceed Parallel convolutional code encoding of the path data to obtain Path check sequence For the first The processing of the path data is as follows: Step 904a, the multiplexer will... and Merge into one data stream, denoted as .

[0072] Step 904b, the convolutional encoder uses all Any path input Road input ,the remaining Road input Output the check subsequence .

[0073] Step 905, the multiplexer will Merge them into one path, denoted as the check sequence. .

[0074] For any parity sequence ,when hour, Initialize to 0. Similarly, for any data sequence... ,when hour, Initialize to 0. In this coupled structure, the above encoding process must satisfy... ,and , All are greater than 0. Assume The length is The length of each input channel of the encoder is , verification sequence The length is The total bitrate of the system is: .

[0075] The core coding unit consists of multiple convolutional encoders, and the parallelism of the encoders can be flexibly adjusted. Adjusting the parallelism has minimal impact on error correction performance.

[0076] Example 4: PSCBCC Encoding and Decoding at 1 / 3 Code Rate First, the encoding parameters are set. The original data frame length is LK (L=50, K=1000 bits), which is divided into 2L=100 sub-data groups. Each group has a length of K / 2 = 500 bits; coupling parameters: =1, First specific moment = Current moment +1, =2, the second specific time = the current time + 2; the parallel number of parallel coding units S = 4, the demultiplexer divides the input into 4 subsequences, the length of each subsequence = 500 / 4 = 125 bits (≥50 bits); the convolutional code encoder is a multi-bit input system convolutional code with a code rate of 2 / 3, generating polynomial octal [133, 171]; the interleaver uses block interleaving (25×5 matrix) to shuffle the order of the previous sub-data and the check sequence.

[0077] Taking t=3 as an example, the encoding steps are as follows: Step 1001, Input: Current sub-data (500 bits), previous sub-data (Interleaved), previous check sequence (After interleaving); Step 1002, the demultiplexer divides the three inputs into four subsequences. (i=1~4), each path is 125 bits long; Step 1003, parallel encoding: merging of the i-th path multiplexer and for Convolutional code encoder and Input: Output: Checksum subsequence Step 1004: The multiplexer merges 4 channels. For the current verification sequence (1000 bits); Step 1005, Storage and Output: Retained up to time t+1=4. Retain until time t+2=5; output the codeword. (1500 bits).

[0078] In the decoding parameter settings, the channel model adopts the AWGN channel, and the received signal is y=x+n; the maximum number of iterations of the iteration parameters is 20, and the iteration order is bidirectional decoding from front to back → from back to front; the decoding algorithm adopts the Log-MAP soft-input soft-output algorithm, and the number of parallel operations is S=4.

[0079] Taking t=3 as an example, the decoding steps include: Step 1006, input 8 sets of LLRs: , , , Step 1007, parallel decoding: Divide the LLR into 4 sub-sequences, and input the i-th decoder... , , Output updated external information and , The predicted values; Step 1008, Merging and Iteration: The multiplexer merges the predicted values ​​to obtain... and Update the external information and repeat the iteration. Output the result after the 20th iteration.

[0080] As shown in Figure 9, the bit error rate of PSCBCC is 10 when the signal-to-noise ratio is -5dB. -5 BCC is 5×10 -5 This verifies the performance advantages of PSCBCC; parallel processing reduces encoding time to 1 / 4 of that of serial schemes, meeting the real-time requirements of high-bandwidth scenarios.

[0081] Example 5: The code rate of the multi-rate extended (rate 1 / 2) convolutional code encoder is k / n = 3 / 5 (n = 5, k = 3), k1 = 2k - n = 1, k2 = nk = 2; sub-data length: Bits, length of each input channel Bits, check sequence length Bits; parallelism S=2, after splitting, the length of each subsequence = 1000 / 2 = 500 bits, the encoding and decoding process is the same as in Example 4, the total system code rate (Adjusting parameters can support a 1 / 2 bitrate). For example, a convolutional code encoder with a bitrate of 3 / 4 (k=3, n=4). , The sub-data length K = 1000 bits, the check sequence length V = 2 × 1000 × 1 / 2 = 1000 bits, and the total code rate = 1000 / (1000+1000) = 1 / 2.

[0082] In summary, this invention achieves a balance between high throughput and high reliability in high-bandwidth communication scenarios through the combination of parallel processing and asymmetric coupling, demonstrating significant practical value. By using parallel encoding / decoding units (with an adjustable number of parallel units S), the processing time for a single data set is reduced to 1 / S; for example, when S=4, the processing time is 1 / 4 of that of serial processing, significantly improving transmission efficiency in high-bandwidth scenarios. The asymmetric coupling structure, with different values ​​for m1 and m2, avoids redundant conflicts. At the same code rate, the bit error rate is reduced by more than one order of magnitude compared to traditional BCC, and performance is improved by more than 0.1dB at low signal-to-noise ratios. Supporting k / n code rates allows for flexible application to diverse scenarios in 4G / 5G evolution and 6G. Performance surpasses LTE-turbo and NR-LDPC standard codes, and the parallelism can be dynamically adjusted according to hardware resources (such as FPGA parallel processing capabilities), enabling flexible engineering implementation.

[0083] Furthermore, an exemplary embodiment of the present invention may also provide a computer-readable storage medium storing a computer program. This computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform a parallel stair-curved convolutional code spatially coupled encoding and decoding method according to an exemplary embodiment of the present invention. This computer-readable recording medium is any data storage device capable of storing data read by a computer system. Examples of computer-readable recording media include: read-only memory, random access memory, read-only optical disk, magnetic tape, floppy disk, optical data storage device, and carrier waves (such as data transmission via the Internet through wired or wireless transmission paths).

[0084] Furthermore, an exemplary embodiment of the present invention may also provide a computing device. The computing device includes a processor and a memory. The memory stores a computer program. The computer program is executed by the processor, causing the processor to execute a computer program of a parallel stair-curved convolutional code spatial coupling encoding and decoding method according to an exemplary embodiment of the present invention.

[0085] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings, disclosure, and other materials. In this specification, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple components. A single processor or other unit can implement several functions listed in the specification. While certain measures are described in different embodiments, this does not mean that these measures cannot be combined to produce good results.

[0086] Although the invention has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made therein without departing from the spirit and scope of the invention. Accordingly, this specification and drawings are merely illustrative of the invention and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is also intended to include such modifications and modifications.

Claims

1. A parallel staircase convolutional code spatial coupling coding method, characterized in that, Includes the following steps: A frame of data is divided into multiple sub-data sequences. For each sub-data sequence, encoding is performed by a parallel coding unit: the sub-data sequence is used as the first input, the previously specific sub-data sequence is interleaved and used as the second input, and the previously specific parity sequence is interleaved and used as the third input to generate the current parity sequence. The parallel coding unit divides the input into S sub-sequences using a demultiplexer, encodes them using S parallel convolutional codes, and then merges them into the current parity sequence using a multiplexer, where S is the number of parallel operations and S≥1. The output contains codewords for the sub-data sequence and the current parity sequence. The sub-data sequence is retained until a subsequent first specific time point as the input to the parallel coding unit, and the current parity sequence is retained until a subsequent second specific time point as the input to the parallel coding unit, with the first and second specific time points being different. If the previously specific sub-data sequence or parity sequence does not exist, it is replaced with an initial value.

2. The parallel staircase convolutional code spatial coupling coding method according to claim 1, characterized in that, The number of the multiple sub-data sequences is twice the number of the original data frame groups; the length of each sub-data group is half the length of a single data group in the original data frame; the length of the current verification sequence is the same as the length of a single data group in the original data frame; the total system code rate is 1 / 3; the value of the parallel number S satisfies that the length of the sub-sequence after each sub-data group is divided is greater than or equal to 50 bits.

3. The parallel staircase convolutional code spatial coupling coding method according to claim 1, characterized in that, The first specific time point is the current time point plus m-1, and the second specific time point is the current time point plus m, where m is an integer greater than 2; the convolutional code encoder in the parallel coding unit is a multi-bit input system convolutional code encoder with a code rate of 2 / 3.

4. The parallel staircase convolutional code spatial coupling coding method according to claim 1, characterized in that, The parallel coding unit employs a multi-bit input system convolutional code encoder with a code rate of k / n, where n is the number of output bits, k is the number of input bits, and nk > 1 and 2k - n ≥ 1. Specifically: the length of each sub-data set is K; the first and second input paths are merged into a k1 = 2k - n sub-sequence; the third input path is a k2 = nk sub-sequence; the length of the current verification sequence is 2Kk2 / k1; and the total system code rate is k1 / (k1 + 2k2).

5. A parallel staircase convolutional code spatial coupling decoding method, used to decode codewords generated by the encoding method as described in claim 1, characterized in that, Includes the following steps: Iterative decoding is performed on the codeword. For the codeword corresponding to each group of sub-data, the channel log-likelihood ratio and extrinsic information of the group of sub-data and the previous specific sub-data, as well as the channel log-likelihood ratio and extrinsic information of the group of parity sequences and the previous specific parity sequences, are input through the parallel decoding unit. The parallel decoding unit divides the input log-likelihood ratio into S-way sub-sequences through a demultiplexer. After decoding by S-way parallel convolutional codes, the sub-sequences are merged into updated extrinsic information by a multiplexer. Output the updated external information and the predicted values ​​of the corresponding sub-data; decode according to the preset iteration order, and output the final decoding result when all codewords are successfully decoded or the maximum number of iterations is reached.

6. The parallel staircase convolutional code spatial coupling decoding method according to claim 5, characterized in that, The input log-likelihood ratio includes 8 groups, namely: channel log-likelihood ratio of the current sub-data, channel log-likelihood ratio of the previous specific sub-data, extrinsic information of the current sub-data, extrinsic information of the previous specific sub-data, channel log-likelihood ratio of the current check sequence, channel log-likelihood ratio of the previous specific check sequence, extrinsic information of the current check sequence, and extrinsic information of the previous specific check sequence; wherein, the initial value of the extrinsic information is 0.

7. The parallel staircase convolutional code spatial coupling decoding method according to claim 5, characterized in that, The preset iteration order includes at least one of the following: sequential decoding from front to back, sequential decoding from back to front, parallel decoding, and sliding window decoding.

8. The parallel staircase convolutional code spatial coupling decoding method according to claim 5, characterized in that, The convolutional code decoder in the parallel decoding unit employs a soft-input soft-output algorithm, including the BCJR algorithm, the MAP algorithm, or the Log-MAP algorithm.

9. A parallel staircase convolutional code spatially coupled coding apparatus for performing the coding method as described in any one of claims 1 to 4, characterized in that, include: The data segmentation module is used to segment a frame of data into multiple sets of sub-data sequences; the parallel encoding unit includes a demultiplexer, an S-way convolutional code encoder and a multiplexer. The demultiplexer segments the current sub-data, the previous specific sub-data, and the previous specific check sequence into S-way sub-sequences. The convolutional code encoder encodes the sub-sequences in parallel. The multiplexer merges the encoding results to generate the current check sequence. The output module is used to output the codeword containing the current sub-data and the current check sequence; The storage module is used to retain the current sub-data until a subsequent first specific time and the current verification sequence until a subsequent second specific time, wherein the first and second specific times are different; and to provide an initial value when the previous specific sub-data or verification sequence does not exist.

10. A parallel staircase convolutional code spatially coupled decoding apparatus for performing the decoding method as described in any one of claims 5 to 8, characterized in that, include: The iteration control module is used to control the iterative decoding process of codewords and set the preset iteration order; The information input module is used to input the channel log-likelihood ratio and extrinsic information of the sub-data and the verification sequence; The parallel decoding unit includes a demultiplexer, an S-way convolutional code decoder, and a multiplexer. The demultiplexer divides the input log-likelihood ratio into S-way subsequences. The decoder decodes the subsequences in parallel. The multiplexer merges the decoding results to generate updated extrinsic information and sub-data prediction values. The result output module is used to output the final decoding result when all codewords are successfully decoded or the maximum number of iterations is reached.

Images