A Sliding Window Soft Serial Cancellation List Decoding Method for Polar Code Modulation

Through the sliding window soft serial cancellation list decoding method, combined with the spatial coupling structure and iterative decoding of soft input and soft output, the shortcomings of the hard output decoding method in the existing technology are solved, the error rate performance is optimized and the decoding is reduced. Complexity.

CN114584152BActive Publication Date: 2024-11-05SOUTHEAST UNIV +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202210215907.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-07
Publication Date
2024-11-05
Estimated Expiration
2042-03-07

AI Technical Summary

Technical Problem

The existing polarized coding encoding modulation technology uses hard output decoding methods on the receiving end, and cannot take advantage of the performance advantages of iterative decoding of soft input and soft output, and the existing soft output decoding methods are not suitable for asynchronous based on spatial coupling structures. Polarized code encoding modulation.

Method used

A sliding window soft serial offset list decoding method is proposed, combining spatial coupling structure and soft input and soft output iterative decoding, and through sliding window iterative and new soft information, the error group rate performance is optimized, and cyclic redundancy verification is used in advance Terminate the iteration process.

Benefits of technology

The system error rate performance is improved, the high coding complexity generated by iteration is reduced, and the decoding complexity under the working signal-to-noise ratio consistent with the traditional method is achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114584152B_ABST
    Figure CN114584152B_ABST
Patent Text Reader

Abstract

The present invention discloses a sliding window soft serial offset list decoding method suitable for polar code modulation, comprising the following steps: initializing a receiving end; t Perform sliding window decoding; if the decoding state s0 = 0, perform the second iteration process; output θ t If s0=1, reconstruct #imgabs1# to get w t The estimate #imgabs2# updates #imgabs3# so that #imgabs4# is consistent with #imgabs5#; let t′=t+L w , calculate #imgabs6# and update s; set t=t+1, return to the second step until the final estimated sequence of all sent source information bit sequences is output. This method combines the advantages of spatial coupling structure and soft-input soft-output iterative decoding to optimize the system's error rate performance.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of coding modulation in wireless communication, and in particular relates to a sliding window soft serial cancellation list decoding method suitable for polar code modulation. Background Art

[0002] The coded modulation technology based on polarization code combines polarization code and high-order modulation, thereby improving the spectrum efficiency of the communication system. The coded modulation technology based on polarization code can be divided into two categories, synchronous modulation and asynchronous modulation. One synchronous modulation technology based on polarization code is multilevel polar-coded modulation (MLPCM). m , m=1,2,3,4,... binary digital modulation, MLPCM divides the input data stream into m priorities, and decodes it in stages according to the priority of the data stream at the receiving end, and achieves the channel capacity through set decomposition mapping. Another synchronous modulation technology is bit-interleaved polar-coded modulation (BIPCM). In BIPCM, each data stream has no priority, so there is only one encoder and decoder at the sending and receiving ends, which has lower processing delay than MLPCM. However, in order to reduce interference between data streams, BIPCM uses Gray mapping, which results in a large mutual information loss, and the error rate performance is inferior to MLPCM.

[0003] On the other hand, if the data streams are superimposed into a spatial coupling structure, BIPCM can still use set decomposition mapping without losing capacity under parallel modulation partitioning. This polar code coded modulation technology is asynchronous polar code coded modulation (A-PCM). At present, A-PCM uses traditional hard-output polar code decoding methods at the receiving end, such as successive cancellation list (SCL) or cyclic redundancy check assisted SCL (CA-SCL) decoding methods. Studies have shown that under the CA-SCL decoding method, A-PCM can obtain better polarization effect and error rate performance than MLPCM. However, CA-SCL outputs hard bit information and cannot take advantage of the performance advantages of soft input and soft output iterative decoding. Existing polar code soft output decoding methods, such as soft list or soft cancellation decoding schemes, are mainly based on synchronous BIPCM and are not suitable for A-PCM based on spatial coupling structure. Summary of the invention

[0004] Technical problem: In order to solve the shortcomings of the above-mentioned background technology, the present invention proposes a sliding window soft serial cancellation list decoding (SWS-SCL) method suitable for polar code modulation. The method combines the advantages of spatial coupling structure and soft-input soft-output iterative decoding, optimizes the system's error rate performance, and is suitable for practical communication applications.

[0005] Technical solution: In order to achieve the above-mentioned invention object, a sliding window soft serial cancellation list decoding method applicable to polar code modulation of the present invention comprises the following steps:

[0006] Step 1: Initialize the receiving end:

[0007] m=1,2,3,4,... represents the number of bits mapped to modulation symbols, and the frame length is N=2 λ ,λ is the power,λ=1,2,...;θ t represents the source information bit sequence of length K at time t, t=1,2,..., Represents θ t Join K pc The length after cyclic redundancy check bits is K+K pc The information bit sequence of the polar code encoder is u t Information Bit and frozen bits, with a length of Indicates that the value is an integer not less than log2m; w t The code word v is the code word after the polar code is encoded by asynchronous polar code modulation. t The sequence of size mN is generated by interleaving the m subsequences of where i = 1, 2, ..., m, Π i (·) is the i-th interleaver, v t subsequence By v t The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i N}, the interleaved subsequence w t,i By t The elements indexed by ε∈{(i-1)N+1,(i-1)N+2,...,iN} in the is the i-th element of the set P, with size m, where P represents the channel mapping mode, which is determined by the equal-capacity channel partitioning mapping method;

[0008] vector Indicates w tThe prior information of length is mN; for any t=1,2,..., Initialize to a sequence of all 0s;

[0009] In the right When performing sliding window decoding, the vector Indicates the current decoding window state, where L w =m+1 is the window size, s l ∈{0,1} represents the decoder pair The decoding state, l = 0, 1, ..., L w :s l =1 indicates that the decoding result passes the cyclic redundancy check; s l =0 means that decoding has not been performed or the decoding result has not passed the cyclic redundancy check;

[0010] Initialize t = 1, and when t = 1 Initialize to an all-zero sequence;

[0011] Step 2: θ t Perform sliding window decoding:

[0012] The maximum number of iterations for each sliding window is N. iter , N iter =2,3,..., the value is determined according to the simulation results; for all n iter ∈{1,2,...,N iter}Perform iterative decoding operation: Each iteration uses the decoder input information to decode L w Information bit sequence Perform cyclic redundancy check to assist serial offset list decoding and obtain Estimates And update the decoding window status according to the cyclic redundancy check result and the input information of the demodulator, and then the demodulator outputs the updated input information of the decoder for the next iteration until the iteration termination condition is met or the maximum number of iterations is reached;

[0013] Step 3: If s0 = 0, perform a second iteration process by adjusting the prior information of the demodulator input Re-align Decode to get

[0014] Step 4: Output

[0015] Step 5: If s0=1, Reconstruct the code to get w t Estimates Push to update Make and consistent, that is

[0016]

[0017] Among them, sign(·) represents the sign function;

[0018] Updated Will be used for θ t+1 Sliding window decoding of

[0019] Step 6: Let t′=t+L w ,calculate

[0020] Step 7: Updates: in When l=0,1,...L w-1 hour, Then The updated s will be used for θ t+1 Sliding window decoding of

[0021] Step 8: Let t=t+1, and return to step 2 until the final estimated sequence of all sent source information bit sequences is output.

[0022] Furthermore, the second step specifically includes the following steps:

[0023] Step 2.1: If Then terminate the second step loop;

[0024] Step 2.2: According to the modulation mode of asynchronous polar code coded modulation, at time t=m,m+1,..., sequence c t =[w t-m+1,1 ,...,w t-m+i,i ,...,w t,m ] is mapped into modulation symbol frame x after set decomposition t , where w t-m+i,i Corresponding to x t The i-th mapping index bit sequence; at time t = 1, 2, ..., m-1, c t =[0 1×(m-t)N ,w 1,m-t+1 ,...,w τ,m-t+τ ,...,w t,m ] is mapped to x through set decomposition t , where τ=1,2,...t,0 1×(m-t)N is a sequence of all zeros of size (mt)N, w τ,m-t+τ Corresponding to x t The m-t+τth mapping index bit sequence; for t′∈{t,t+1,...,t+Lw -1}, perform the following operations in sequence:

[0025] Step 2.2.1. For all i∈{1,2,...,m}, calculate the modulation symbol frame x t′+m-i The i-th mapping index bit sequence w in t′,i External information

[0026] Step 2.2.2: Calculate the codeword v after polar code encoding t′ Prior information

[0027] Step 2.2.3: As the input information of the decoder, the serial cancellation list decoding method based on log-likelihood ratio is used to output v t′ J estimated paths and their path metrics β1,β2,...,β j ,..,β J , after cyclic redundancy check, we get θ t′ Estimates And update s according to the verification results t′-t ;

[0028] Step 2.2.4, calculate v t′ The posterior information and external information

[0029] Step 2.2.5: Update the demodulator input information

[0030] Further, step 2.2.1 The calculation process is as follows:

[0031]

[0032] in, subsequence Depend on The index in is ε∈{(i-1)N+1,(i-1)N+2,...,iN}, corresponding to w t′,i Prior information of Represents the modulation symbol frame x t′+m-i The i-th mapping index bit sequence w in t′,i The posterior information of The nth element of It is given by the following formula, n = 1, 2, ..., N:

[0033]

[0034] in, and Serves 2 m A subset of the binary modulation symbol set X, for any symbol The i-th mapped bit of x For any symbol [q t′+m-i,i ] n is a vector q of length N t′+m-i,i The nth element of is calculated as:

[0035]

[0036] Among them, σ 2 represents the variance of the complex Gaussian noise sequence; [h t′+m-i ] n and [y t′+m-i ] n They represent the channel gain sequence h at time t′+mi respectively. t′+m-i and the received signal y t′+m-i The nth element of t′+i′-i,i′ is an intermediate variable, calculated by the following formula:

[0037]

[0038] in, for The nth element of .

[0039] Furthermore, the codeword v after polar code encoding is calculated in step 2.2.2 t′ Prior information The process is as follows:

[0040] For any and The value of the element with index δ∈{(p-1)N+1,(p-1)N+2,...,pN} is 0; The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i The elements of N} are For all i∈{1,2,...,m}, calculate as follows

[0041]

[0042] in, Yes i The inverse function of (·);

[0043] So as to obtain the updated

[0044] Further, step 2.2.4 The calculation process is as follows:

[0045] The kth element of Calculate as follows:

[0046]

[0047] Among them, the collection gather represents J estimated paths The path index set whose kth element is 0, represents J estimated paths The path index set whose kth element is 1, yes The kth element of ;

[0048] Said The calculation formula is as follows:

[0049]

[0050] Further, step 2.2.5 The calculation process is as follows:

[0051] The element with index ε∈{(i-1)N+1,(i-1)N+2,...,iN} in is For all i∈{1,2,...,m}, calculate as follows

[0052]

[0053] in, Depend on The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i N} element composition;

[0054] So as to obtain the updated

[0055] Furthermore, the third step specifically includes the following process:

[0056] Step 3.1. Setup All are zero sequences;

[0057] Step 3.2: If Then terminate the third step loop;

[0058] Step 3.3: For t′∈{t,t+1,...,t+L w -1}, follow steps 2.2.1 to 2.2.5 in sequence.

[0059] Furthermore, the sixth step specifically includes the following process:

[0060] Step 6.1. For all i∈{1,2,...,m}, calculate

[0061] Step 6.2: Calculate v t′ Prior information

[0062] Step 6.3: As the input of the decoder, the cyclic redundancy check-assisted serial cancellation list decoding method based on log-likelihood ratio is used to obtain θ t′ Estimates And update according to the verification results

[0063] Beneficial effects: Compared with the prior art, the present invention has the following beneficial effects: Compared with CA-SCL and SCL, the present invention combines the spatial coupling structure and soft-input soft-output iterative decoding, which can further optimize the error rate performance. In addition, the present invention can terminate the iterative process in advance through the cyclic redundancy check, and the decoding complexity under the system working signal-to-noise ratio is consistent with that of CA-SCL or SCL. Specific embodiments

[0064] The present invention will be further described below.

[0065] The present invention proposes a sliding window soft serial cancellation list decoding method suitable for polar code modulation. On the one hand, the soft information output by the path metric calculation of the SCL decoding method is used. On the other hand, the spatial coupling structure of A-PCM is used to continuously iterate and update the soft information using the sliding window decoding method, thereby maximizing the system error rate performance. In addition, the present invention also uses a cyclic redundancy check method in the iterative process. According to the check result, the early termination condition of the iteration is set, thereby reducing the high decoding complexity generated by the iteration.

[0066] The transmitter is described as follows: Let m = 1, 2, 3, 4, ..., which means the number of bits mapped to modulation symbols, that is, the symbol modulation order is 2 m , frame length is N = 2 λ ,λ is the power,λ=1,2,.... θ t Represents a source information bit sequence with a length of K at time t, t=1, 2, ... Represents θ t Join K pc The length after cyclic redundancy check bits is (K+K pc ) information sequence.

[0067] The input bit sequence u of the polar code encoder is t Information Bit and frozen bits, with a length of Indicates that the value is an integer not less than log2m. in u t The index in is set A, that is, A is a set The subset of u is determined by the Gaussian approximation method. t The remaining bits are frozen bits and are set to 0.

[0068] According to the polar code encoding method, the encoded codeword in is a two-dimensional binary kernel matrix, yes The generator matrix of order, is the Kronecker product. A-PCM contains m interleavers that respectively t Interleave the m subsequences of t , expressed as where i = 1, 2, ..., m, Π i (·) is the i-th interleaver, is the i-th element of the set P, with a size of m, where P represents the channel mapping mode, which is determined by the equal-capacity channel segmentation mapping method; By v t The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i N}; the interleaved subsequence w t,i By t The elements in which the index is ε∈{(i-1)N+1,(i-1)N+2,...,iN}. t The bits that do not participate in the interleaver, that is, for any and v t The elements with indices δ∈{(p-1)N+1,(p-1)N+2,...,pN} in are punctured bits. The length of the punctured bit sequence is According to the A-PCM modulation method, at t=m,m+1,..., the sequence c t =[w t-m+1,1 ,...,w t-m+i,i,...,w t,m ] is mapped into modulation symbol frame x after set decomposition t , where w t-m+i,i Corresponding to x t The i-th mapping index bit sequence; at time t = 1, 2, ..., m-1, c t =[0 1×(m-t)N ,w 1,m-t+1 ,...,w τ,m-t+τ ,...,w t,m ] is mapped to x through set decomposition t , where τ=1,2,...t,0 1×(m-t)N is a sequence of all zeros of size (mt)N, w τ,m-t+τ Corresponding to x t The (m-t+τ)th mapping index bit sequence.

[0069] With N=128, K=256, hexadecimal quadrature amplitude modulation, L w =5,N iter =15 is taken as an example to illustrate the SWS-SCL decoding method of the present invention.

[0070] Step 1: Initialize the receiving end. Let vector Indicates w t The length of the prior information is mN. For any t=1,2,..., Initialized to a sequence of all 0s. When performing sliding window decoding, the vector Indicates the current decoding window state, where L w =m+1 is the window size, s l ∈{0,1},l=0,1,...,L w Decoder pair Decoding status: s l =1 indicates that the decoding result passes the cyclic redundancy check; s l =0 means that decoding has not been performed or the decoding result has not passed the cyclic redundancy check. Initialize t=1, and set t=1 Initialized to a sequence of all zeros.

[0071] The above steps are explained in detail below: In this example, m=4, A-PCM contains 4 interleavers. By the equal capacity channel segmentation mapping method, the channel mapping pattern P = {1, 3, 2, 4} is obtained. t All bits in the modulation process participate in the modulation process, where v 1,1 =w 1,1 ,v 1,3 =w 1,2 ,v 1,2 =w1,3 ,v 1,4 =w 1,4 ,. According to the A-PCM modulation method, at t = 1, c1 = [0 1×3N ,w 1,4 ]; at t = 2, c2 = [0 1×2N ,w 1,3 ,w 2,4 ]; at t = 3, c3 = [0 1×N ,w 1,2 ,w 2,3 ,w 3,4 ]; at t=4,5,..., sequence c t =[w t-3,1 ,w t-2,2 ,w t-1,3 ,w t,4 ]. t =[0 1×(m-t)N ,w 1,m-t+1 ,...,w τ,m-t+τ ,...,w t,m ] is mapped to x through set decomposition t .vector Initialized to an all-zero sequence to indicate that the decoder has not yet Decode.

[0072] Step 2: θ t Perform sliding window decoding.

[0073] The maximum number of iterations for each sliding window is N. iter , N iter =2,3,..., its value is determined according to the simulation results. Each iteration uses the input information of the decoder to w Information bit sequence Perform CA-SCL decoding and get Estimates And update the decoding window status according to the cyclic redundancy check result The demodulator then outputs the updated decoder input information for the next iteration until the iteration termination condition is met or the maximum number of iterations is reached. iter ∈{1,2,...,N iter}Perform iterative decoding operation. This step includes the following process:

[0074] Step 2.1: If Then terminate the second step loop.

[0075] Step 2.2: For t′∈{t,t+1,...,t+L w -1}, perform the following operations in sequence.

[0076] Step 2.2.1, let vector and Respectively represent x t′+m-i The i-th mapping index bit sequence w in t′,i External information and posterior information; for all i∈{1,2,...,m}, calculate

[0077]

[0078] Neutron sequence Depend on The index in is ε∈{(i-1)N+1,(i-1)N+2,...,iN}, corresponding to w t′,i prior information. The nth, n=1,2,...,N,element of Given by the following formula

[0079]

[0080] In the above formula and Serves 2 m A subset of the binary modulation symbol set X. For any symbol The i-th mapped bit of x For any symbol [q t′+m-i,i ] n is a vector q of length N t′+m-i,i The nth element of is calculated as

[0081]

[0082] In the above formula, |·| represents the modulo function, σ 2 represents the complex Gaussian noise sequence z t The variance of t′+m-i ] n and [y t′+m-i ] n They represent the channel gain sequence h at time (t′+mi) respectively. t′+m-i and the received signal y t′+m-i The nth element of t′+i′-i,i′ is an intermediate variable, calculated by the following formula

[0083]

[0084] In the above formula for The nth element of .

[0085] Step 2.2.2, calculate v t′ Prior information

[0086] For any and The element indexed by δ∈{(p-1)N+1,(p-1)N+2,...,pN} in the sequence is 0; express The prior information of The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i N}. For all i∈{1,2,...,m}, calculate

[0087]

[0088] In the above formula Yes i The inverse function of (·).

[0089] So as to obtain the updated

[0090] Step 2.2.3: As the input of the decoder, the SCL decoding method based on log-likelihood ratio is used to output v t′ J estimated paths and their path metrics β1,β2,...,β j ,..,β J , after cyclic redundancy check, we get θ t′ Estimates And update st′-t according to the verification result.

[0091] Step 2.2.4, calculate v t′ The posterior information and external information

[0092]

[0093] in yes The kth element of the set gather in is a bit sequence The kth element of represents J estimated paths The path index set whose kth element is 0, represents J estimated paths The path index set whose k-th element is 1.

[0094] The calculation formula is as follows:

[0095]

[0096] Step 2.2.5, Update

[0097] For all i=1,2,...,m, update using the following formula

[0098]

[0099] Then make The element with index ε∈{(i-1)N+1,(i-1)N+2,...,iN} in is So as to obtain the updated

[0100] Step 3: If s0 = 0, perform the second iteration process. This step includes the following process.

[0101] Step 3.1. Setup All are 0 sequences.

[0102] Step 3.2: If Then terminate the third step loop.

[0103] Step 3.3: For t′∈{t,t+1,...,t+L w -1}, perform the following operations in sequence.

[0104] Step 3.3.1. For all i∈{1,2,...,m}, use formulas (1)-(4) to calculate

[0105] Step 3.3.2, calculate v t′ Prior information For any and The elements with index δ∈{(p-1)N+1,(p-1)N+2,...,pN} in ∈{1,2,...,m} are taken to be 0. For all i∈{1,2,...,m}, use formula (5) to calculate Then make The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i The elements of N} are So as to obtain the updated

[0106] Step 3.3.3: As the input of the decoder, the SCL decoding method based on log-likelihood ratio is used to output v t′ J estimated paths and their path metrics β1,β2,...,β j ,..,β J , after cyclic redundancy check, we get θ t′ Estimates And update s according to the verification results t′-t .

[0107] Step 3.3.4. Calculate v using formulas (6) and (7) t′ The posterior information and external information

[0108] Step 3.3.5: For all i=1,2,...,m, update using formula (8)

[0109] Step 4: Output

[0110] The above is explained in detail below: From the second step, the maximum number of iterations for each sliding window is 15, and each iteration counts L w Information bit sequence In the second step, step 2.1 sets the termination condition of the iterative process, that is, if the decoder is The decoding of all passed the cyclic redundancy check. Then the loop of the second step is terminated. This step can reduce the high decoding complexity caused by multiple iterations by terminating the iteration process in advance. Perform CA-SCL decoding and update the decoding status according to the cyclic redundancy check result Step 2.2.1 and step 2.2.2 update the decoder input information Steps 2.2.4 and 2.2.5 update the demodulator input information based on the SCL decoder output When the first iteration, i.e. the second step, decodes After failure (i.e., s0=0), the third step of the second iteration adjusts the demodulator's prior information Re-align Decode. The fourth step is to output the final decoding result.

[0111] Step 5: If s0=1, Reconstruct the code to get w t Estimates renew Make and consistent, that is

[0112]

[0113] In the above formula, sign(·) represents a sign function.

[0114] Updated Will be used for θ t+1 Sliding window decoding.

[0115] Step 6: Let t′=t+L w ,calculate This step includes the following processes:

[0116] Step 6.1. Use formulas (1)-(4) to calculate for all i∈{1,2,...,m}

[0117] Step 6.2: Calculate v t′ Prior information For any and The elements with index δ∈{(p-1)N+1,(p-1)N+2,...,pN} in ∈{1,2,...,m} are taken to be 0. For all i∈{1,2,...,m}, use formula (5) to calculate Then make The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i The elements of N} are So as to obtain the updated

[0118] Step 6.3: As the input of the decoder, the CA-SCL decoding method based on log-likelihood ratio is used to obtain θ t′ Estimates And update according to the verification results

[0119] Step 7: Update s. Let in When l=0,1,...L w-1 hour, Then make The updated s will be used for θ t+1 Sliding window decoding.

[0120] Step 8: Let t=t+1, and return to step 2 until the final estimated sequence of all sent source information bit sequences is output.

[0121] The above steps are explained in detail below: Steps 5 to 7 mainly update the iteration information and decoding state s for the next decoding window pair θ t+1 until all the sent source information bit sequences are decoded.

[0122] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A sliding window soft serial cancellation list decoding method suitable for polar code modulation, characterized in that: The following steps are involved: Step 1: Initialize the receiving end: m=1,2,3,4,... represents the number of bits mapped to modulation symbols, and the frame length is N=2 λ ,λ is the power,λ=1,2,...;θ t represents the source information bit sequence of length K at time t, t=1,2,..., Represents θ t Join K pc The length after cyclic redundancy check bits is K+K pc The information bit sequence of the polar code encoder is u t Information Bit and frozen bits, with a length of Indicates that the value is an integer not less than log2m; w t The code word v is the code word after the polar code is encoded by asynchronous polar code modulation. t The sequence of size mN is generated by interleaving the m subsequences of where i = 1, 2, ..., m, Π i (·) is the i-th interleaver, v t subsequence By v t The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i N}, the interleaved subsequence w t,i By t The elements indexed by ε∈{(i-1)N+1,(i-1)N+2,...,iN} in the is the i-th element of the set P, with size m, where P represents the channel mapping mode, which is determined by the equal-capacity channel partitioning mapping method; vector Indicates w t The prior information of length is mN; for any t=1,2,..., Initialize to a sequence of all 0s; In the right When performing sliding window decoding, the vector Indicates the current decoding window state, where L w =m+1 is the window size, s l ∈{0,1} represents the decoder pair The decoding state, l = 0, 1, ..., L w :s l =1 indicates that the decoding result passes the cyclic redundancy check; s l =0 means that decoding has not been performed or the decoding result has not passed the cyclic redundancy check; Initialize t = 1, and when t = 1 Initialize to an all-zero sequence; Step 2: θ t Perform sliding window decoding: The maximum number of iterations for each sliding window is N. iter , N iter =2,3,..., the value is determined according to the simulation results; for all n iter ∈{1,2,...,N iter }Perform iterative decoding operation: Each iteration uses the decoder input information to decode L w Information bit sequence Perform cyclic redundancy check to assist serial offset list decoding and obtain Estimates And update the decoding window status according to the cyclic redundancy check result and the input information of the demodulator, and then the demodulator outputs the updated input information of the decoder for the next iteration until the iteration termination condition is met or the maximum number of iterations is reached; Step 3: If s0 = 0, perform a second iteration process by adjusting the prior information of the demodulator input Re-align Decode to get Step 4: Output Step 5: If s0=1, Reconstruct the code to get w t Estimates Push to update Make and consistent, that is Among them, sign(·) represents the sign function; Updated Will be used for θ t+1 Sliding window decoding of Step 6: Let t′=t+L w ,calculate Step 7: Updates: in When l=0,1,...L w-1 hour, Then The updated s will be used for θ t+1 Sliding window decoding of Step 8: Let t=t+1, and return to step 2 until the final estimated sequence of all sent source information bit sequences is output.

2. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 1, characterized in that: The second step specifically includes the following steps: Step 2.1: If Then terminate the second step loop; Step 2.2: According to the modulation mode of asynchronous polar code coded modulation, at time t=m,m+1,..., sequence c t =[w t-m+1,1 ,...,w t-m+i,i ,...,w t,m ] is mapped into modulation symbol frame x after set decomposition t , where w t-m+i,i Corresponding to x t The i-th mapping index bit sequence; at time t = 1, 2, ..., m-1, c t =[0 1×(m-t)N ,w 1,m-t+1 ,...,w τ,m-t+τ ,...,w t,m ] is mapped to x through set decomposition t , where τ=1,2,...t,0 1×(m-t)N is a sequence of all zeros of size (mt)N, w τ,m-t+τ Corresponding to x t The m-t+τth mapping index bit sequence; for t′∈{t,t+1,...,t+L w -1}, perform the following operations in sequence: Step 2.2.

1. For all i∈{1,2,...,m}, calculate the modulation symbol frame x t′+m-i The i-th mapping index bit sequence w in t′,i External information Step 2.2.2: Calculate the codeword v after polar code encoding t′ Prior information Step 2.2.3: As the input information of the decoder, the serial cancellation list decoding method based on log-likelihood ratio is used to output v t′ J estimated paths and their path metrics β1,β2,...,β j ,..,β J , after cyclic redundancy check, we get θ t′ Estimates And update st′-t according to the verification result; Step 2.2.4, calculate v t′ The posterior information and external information Step 2.2.5: Update the demodulator input information 3. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 2, characterized in that: Step 2.2.1 The calculation process is as follows: in, subsequence Depend on The index in is ε∈{(i-1)N+1,(i-1)N+2,...,iN}, corresponding to w t′,i Prior information of Represents the modulation symbol frame x t′+m-i The i-th mapping index bit sequence w in t′,i The posterior information of The nth element of It is given by the following formula, n = 1, 2, ..., N: in, and Serves 2 m A subset of the binary modulation symbol set X, for any symbol The i-th mapped bit of x For any symbol [q t′+m-i,i ] n is a vector q of length N t′+m-i,i The nth element of is calculated as: Among them, σ 2 represents the variance of the complex Gaussian noise sequence; [h t′+m-i ] n and [y t′+m-i ] n They represent the channel gain sequence h at time t′+mi respectively. t′+m-i and the received signal y t′+m-i The nth element of t′+i′-i,i′ is an intermediate variable, calculated by the following formula: in, for The nth element of .

4. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 2, characterized in that: Step 2.2.2 calculates the codeword v after polar code encoding t′ Prior information The process is as follows: For any and The value of the element with index δ∈{(p-1)N+1,(p-1)N+2,...,pN} is 0; The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i The elements of N} are For all i∈{1,2,...,m}, calculate as follows in, Yes i The inverse function of (·); So as to obtain the updated 5. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 2, characterized in that: As described in step 2.2.4 The calculation process is as follows: The kth element of Calculate as follows: Among them, the collection gather represents J estimated paths The path index set whose kth element is 0, represents J estimated paths The path index set whose kth element is 1, yes The kth element of ; Said The calculation formula is as follows:

6. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 2, characterized in that: As described in step 2.2.5 The calculation process is as follows: The element with index ε∈{(i-1)N+1,(i-1)N+2,...,iN} in is For all i∈{1,2,...,m}, calculate as follows in, Depend on The index is η∈{(p i -1)N+1,(p i -1)N+2,...,p i N} element composition; So as to obtain the updated 7. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 1, characterized in that: The third step specifically includes the following process: Step 3.

1. Setup All are zero sequences; Step 3.2: If Then terminate the third step loop; Step 3.3: For t′∈{t,t+1,...,t+L w -1}, follow steps 2.2.1 to 2.2.5 in sequence.

8. The sliding window soft serial cancellation list decoding method applicable to polar code modulation according to claim 1, characterized in that: The sixth step specifically includes the following process: Step 6.

1. For all i∈{1,2,...,m}, calculate Step 6.2: Calculate v t′ Prior information Step 6.3: As the input of the decoder, the cyclic redundancy check-assisted serial cancellation list decoding method based on log-likelihood ratio is used to obtain θ t′ Estimates And update according to the verification results

Citation Information

Cited By

  • Soft input soft output decoding method and device suitable for polar code SoSCL decoder

    CN121308765A

  • Soft input soft output decoding method and device suitable for polar code soscl decoder

    CN121308765B