Coding and decoding method of polarization code under interference channel

Abstract

The invention discloses a polar code coding and decoding method under an interference channel, and belongs to the field of channel coding. The method mainly comprises the following steps: constructing an interference channel model; the method comprises the following steps of: determining a reliability index sequence of an equivalent bit sub-channel through an improved polarization weight (PW) method; coding, modulation and interleaving of polarization codes are completed; and decoding is carried out at a receiving end, soft decision decoding based on Log Likelihood Ratio (LLR) is adopted, and an estimated value of a corresponding bit is obtained by judging the LLR. On one hand, the improved PW algorithm provided by the invention can measure the reliability of the equivalent bit sub-channel on the premise of unknown interference conditions, and has relatively low complexity. And on the other hand, in order to avoid information loss caused by a traditional LLR erasing method during decoding of an interference scene, the LLR value of an interfered position is approximately calculated by using a generalized likelihood ratio, and a lower bit error rate is obtained.

Description

Technical Field

[0001] The invention belongs to the field of channel coding and relates to a polarization code encoding method under an interference channel. Background Art

[0002] Communication interference and anti-interference are eternal contradictions in electromagnetic spectrum warfare. As the electromagnetic environment becomes increasingly complex, harsh and changeable, this trend drives the communication anti-interference technology to change from traditional frequency hopping to multi-domain anti-interference in time domain, space domain, frequency domain, code domain, etc. Channel coding, as one of the most important technologies in communication systems, directly affects the transmission performance of the system. The polar code first proposed by Professor Arikan is the first coding scheme that theoretically proves to reach rather than approach the Shannon limit. As a forward error correction (FEC) coding, it has excellent error correction performance and perfect theoretical basis, and is therefore widely used in anti-interference communication systems. The construction of polar codes in interference scenarios mostly uses Gaussian approximation methods or Monte Carlo methods, but these methods are highly complex and rely on the constructed signal-to-noise ratio and the number of Monte Carlo simulations. For the decoding problem under complex interference, the traditional method is to set the log likelihood ratio (LLR) value of the interfered subchannel to 0, that is, to erase the LLR, and then the decoder uses the additional redundant information in the codeword for decoding. However, for the interfered subchannel, the LLR erasure method is a kind of information loss. Therefore, the document "HUSAINB, MOHAMED H, HAERING L, et al. Deep Learning-Based Multi-Tone Interference Suppression for Short Polar Codes [C]. IEEE 93rd Vehicular Technology Conference, Helsinki, Finland, 2021: 1-5." uses the probability of subcarrier interference as the corresponding LLR weighting factor instead of the erased LLR value, and achieves good results. However, this method requires the additional use of a neural network to calculate the probability of subcarrier interference. Summary of the invention

[0003] In view of the above background, the present invention provides a polar code encoding and decoding method under an interference channel. When the interference condition is unknown at the transmitting end, a low-complexity improved PW method is used to construct an equivalent bit subchannel index sequence with high reliability. During decoding, in order to avoid the information loss caused by the traditional LLR erasure method, a generalized likelihood ratio is used to approximate the LLR value of the interfered position.

[0004] The specific technical solutions of the present invention are as follows:

[0005] S1: Construct interference channel model;

[0006] S2: Determine the reliability index sequence of the equivalent bit subchannel by using the improved polarization weights (PW) method;

[0007] S3: complete the encoding, modulation and interleaving of polar codes;

[0008] S4: Decoding at the receiving end uses soft decision decoding based on log likelihood ratio (LLR) information to determine the LLR and obtain an estimated value of the corresponding bit.

[0009] Furthermore, in step S1, the interference channel is modeled as an additive white Gaussian noise (AWGN) channel with partial frequency band interference, which is expressed as

[0010]

[0011] Among them, s i represents the modulation symbol, n i Represents a mean of 0 and a variance of σ 2 Gaussian white noise, w i Represents the interference noise of some frequency bands, with a mean of 0 and a variance of It is simulated by narrow-band Gaussian white noise.

[0012] Furthermore, in step S2, the reliability index sequence of the equivalent bit subchannel is determined by using the improved PW method. Specifically, according to

[0013]

[0014] To calculate each equivalent bit subchannel The polarization weight Where, for a natural number n>2, the subchannel index k, 1≤k≤N, N=2 n , (b k,n-1 ,…,b k,1 ,b k,0 ) represents the binary expansion of the number (k-1), β = 1.1892, N max =1024, N is the code length, K is the number of information bits;

[0015] The calculated polarization weights are then sorted from large to small, and the corresponding first K subchannel indexes are reliable equivalent bit subchannel sequences.

[0016] Furthermore, in step S4, when calculating the LLR value required for decoding, for the subchannel that is not interfered with, its LLR value is calculated according to

[0017]

[0018] to calculate;

[0019] For the interfered subchannel, in order to avoid the information loss caused by the traditional LLR erasure method, the generalized likelihood ratio is used to approximate the LLR value of the interfered position, that is,

[0020]

[0021] Beneficial Effects

[0022] The present invention provides a polar code encoding and decoding method under an interference channel. In the case of unknown interference at the transmitting end, an improved PW algorithm is used to construct an equivalent bit subchannel index sequence with high reliability, which has lower complexity than traditional Gaussian approximation and Monte Carlo methods. At the same time, in order to avoid the information loss caused by the traditional LLR erasure method during decoding of interference scenes, a generalized likelihood ratio is used to approximate the LLR value of the interfered position, thereby obtaining a lower bit error rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for illustration:

[0024] Figure 1 It is a flowchart of the steps of a polar code encoding method under an interference channel provided by the present invention;

[0025] Figure 2 It is a schematic diagram of a traditional encoding and decoding system of a polar code under an interference channel provided by the present invention. DETAILED DESCRIPTION

[0026] In order to better understand the purpose, structure and function of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings.

[0027] The present invention provides a polar code encoding method under interference channel, the specific process is as follows: Figure 1 As shown, the following steps are included:

[0028] S1: Construct an interference channel model. In this embodiment, the interference channel is modeled as an AWGN channel with partial frequency band interference, which is expressed as

[0029]

[0030] Among them, s i represents the modulation symbol, n iRepresents a mean of 0 and a variance of σ 2 Gaussian white noise, w i Represents the interference noise of some frequency bands, with a mean of 0 and a variance of In practice, there are many types of communication interference signals, such as single-tone interference, multi-tone interference, swept frequency interference, noise frequency modulation interference, etc. This embodiment uses partial frequency band interference as the interference source, and researchers in this field are not limited to this.

[0031] S2: Determine the reliability index sequence of the equivalent bit subchannel by using an improved polarization weights (PW) method. Specifically, according to

[0032]

[0033] To calculate each equivalent bit subchannel The polarization weight Where, for a natural number n>2, the subchannel index k, 1≤k≤N, N=2 n , (b k,n-1 ,…,b k,1 ,b k,0 ) represents the binary expansion of the number (k-1), β = 1.1892, N max =1024, N is the code length, K is the number of information bits; then the calculated polarization weights are sorted from large to small, and the corresponding first K subchannel indexes are reliable equivalent bit subchannel sequences. In this embodiment, the code length of the polarization code is N=256, the code rate is 1 / 2, the number of information bits is K=128, and the polarization weight values ​​of 256 equivalent bit subchannels are calculated according to the improved PW method, the polarization weight values ​​are sorted from large to small, and the equivalent bit subchannel indexes corresponding to the first 128 polarization weight values ​​are selected as the information bit set.

[0034] S3: Complete polar code encoding, modulation, and interleaving. In this embodiment, cyclic redundancy check (CRC)-assisted polar code encoding is used, and 16 of the 128 information bits obtained in step S2 are selected as CRC check bits to generate binary information bits with equal probability. The 16-bit CRC check bits are added to the end of the 112-bit information bits, and the codeword is obtained by polar code encoding. After that, BPSK modulation is performed to obtain the modulation symbol In order to improve the anti-interference capability of the codewords, pseudo-random interleaving is introduced in this embodiment to disrupt the order of the codewords and discretize the continuous codeword errors.

[0035] S4: Decoding at the receiving end uses soft decision decoding based on log likelihood ratio (LLR) information to determine the estimated value of the corresponding bit. Figure 2 The figure shows the traditional encoding and decoding method of polar codes under interference channels. When decoding, the LLR of the interfered position is erased, but it will cause certain information loss. In this embodiment, 60 of the 256 transmission sub-channels are subject to partial frequency band interference, and Before decoding, deinterleaving is performed to restore the original codeword order. When calculating the LLR value required for decoding, for the subchannel that is not interfered with, its LLR value is calculated according to

[0036]

[0037]

[0038] For the interfered subchannel, in order to avoid the information loss caused by the traditional LLR erasure method, the generalized likelihood ratio is used to approximate the LLR value of the interfered position, that is,

[0039]

[0040] in They represent P(y i |s i =1) and P(y i |s i =-1), let Substituting into the above formula, we get

[0041]

[0042] After calculating the LLR values ​​of all symbols at the receiving end, the LLR value of each bit symbol at the source end is iteratively calculated based on the butterfly diagram structure of the polar code. In this embodiment, CRC-assisted SCL decoding is used. When the source bit u i The value does not conform to the When making a hard decision, add a penalty value to the path metric Finally, the L paths with the smallest path metric values ​​are selected as candidate decoding sequences, and the candidate sequences that pass the CRC check are used as the final decoding output.

Claims

1. The method according to claim 1, characterized in that: The interference channel is modeled as an additive white Gaussian noise (AWGN) channel with partial frequency band interference and is expressed as Among them, s i represents the modulation symbol, n i Represents a mean of 0 and a variance of σ 2 Gaussian white noise, w i Represents the interference noise of some frequency bands, with a mean of 0 and a variance of It is simulated by narrow-band Gaussian white noise.

2. The method according to claim 2, characterized in that: The reliability index sequence of the equivalent bit subchannel is determined by using the improved PW method. Specifically, according to To calculate each equivalent bit subchannel The polarization weight Where, for a natural number n>2, the subchannel index k, 1≤k≤N, N=2 n , (b k,n-1 ,…,b k,1 ,b k,0 ) represents the binary expansion of the number (k-1), β = 1.1892, N max =1024, N is the code length, K is the number of information bits; The calculated polarization weights are then sorted from large to small, and the corresponding first K subchannel indexes are reliable equivalent bit subchannel sequences.

3. The method according to claim 3, characterized in that: When calculating the LLR value required for decoding, for the subchannel that is not interfered with, its LLR value is calculated according to to calculate; For the interfered subchannel, in order to avoid the information loss caused by the traditional LLR erasure method, the generalized likelihood ratio is used to approximate the LLR value of the interfered position, that is,

Images