CN111200441B - Polar code decoding method, device, equipment and readable storage medium - Google Patents
Polar code decoding method, device, equipment and readable storage medium Download PDFInfo
- Publication number
- CN111200441B CN111200441B CN202010229739.5A CN202010229739A CN111200441B CN 111200441 B CN111200441 B CN 111200441B CN 202010229739 A CN202010229739 A CN 202010229739A CN 111200441 B CN111200441 B CN 111200441B
- Authority
- CN
- China
- Prior art keywords
- decoding
- channel receiving
- receiving sequence
- neural network
- disturbance noise
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/03—Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words
- H03M13/05—Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words using block codes, i.e. a predetermined number of check bits joined to a predetermined number of information bits
- H03M13/13—Linear codes
- H03M13/15—Cyclic codes, i.e. cyclic shifts of codewords produce other codewords, e.g. codes defined by a generator polynomial, Bose-Chaudhuri-Hocquenghem [BCH] codes
Abstract
The invention discloses a Polar code decoding method, a Polar code decoding device, Polar code decoding equipment and a readable storage medium, wherein the method comprises the following steps of: acquiring a channel receiving sequence with decoding failure; generating disturbance noise matched with a channel receiving sequence; correcting likelihood information of a channel receiving sequence by using the disturbance noise; and decoding the channel receiving sequence by using the corrected likelihood information. In the method, the channel receiving sequence with the decoding failure can be shifted to an error correction area based on the stochastic resonance phenomenon by generating the disturbance noise corresponding to the channel receiving sequence with the decoding failure. That is, after the disturbance noise is obtained, the likelihood information of the corresponding channel receiving sequence can be corrected by using the disturbance noise, and then the channel receiving sequence is decoded by following the likelihood information. Therefore, in the method, even if the decoding fails, the decoding can be carried out by generating the corresponding disturbance noise, so that the decoding success rate can be improved.
Description
Technical Field
The present invention relates to the field of communications technologies, and in particular, to a Polar code decoding method, apparatus, device, and readable storage medium.
Background
Polar code is a channel coding based on channel polarization theory, and a Successive Cancellation (SC) decoding algorithm is a decoding algorithm unique to Polar code provided aiming at the structure of Polar code. Under the SC algorithm, the Polar code is obtained through strict mathematical demonstration, and error-free transmission can be carried out on the B-DMC channel. Meanwhile, the SC decoding algorithm has a low computational complexity, which is only o (nlogn). However, in practical situations, the code length cannot be "long enough", and once an erroneous bit decision occurs, due to the nature of sequential decoding, the erroneous bit has no chance to be corrected, so the decoding performance of the SC decoding algorithm is poor.
Therefore, aiming at medium and short code lengths, an improved algorithm based on the SC decoding algorithm becomes a research hotspot of Polar code decoding, and a series of algorithms are proposed by a plurality of scholars to improve the SC decoding performance of Polar codes. SC-list (scl) coding techniques are capable of obtaining the best maximum likelihood (ml) performance with a sufficiently large list size. Although SCL decoding has superior performance, it has high computational complexity, large memory requirements, and low decoding throughput. Recent scholars have proposed SC-Stack (SCs) decoding algorithms with computational complexity very close to SC decoding algorithms at high signal-to-noise ratio (SNR), and with lower decoding complexity than SCL decoding. Meanwhile, an improved Polar code path metric and an effective stack decoding algorithm are provided, SCS decoding needs larger space complexity, and when the size of a stack is smaller, the performance is slightly poor. Another low complexity solution is SC-flip (SCF) decoding, where an SC decoding failure (detected by CRC check) attempts to identify and flip the first bit error that occurred during the previous decoding, further attempts until the decoding passes CRC check, or the maximum number of attempts is reached. Appropriately combining Fano sequential decoding into standard SC decoding, an alternative improvement of SC decoding is proposed, which decides at each decoding stage whether to move forward or backward along the current path to find more likelihood paths, which may overcome the main drawbacks of SC decoding. However, due to the lack of a complete mathematical characterization, the optimal flipping strategy (flipping bit) remains an open problem.
In summary, how to effectively improve the decoding performance of Polar codes is a technical problem that those skilled in the art are urgently required to solve.
Disclosure of Invention
The invention aims to provide a Polar code decoding method, a Polar code decoding device, Polar code decoding equipment and a readable storage medium, which can improve the decoding performance of Polar codes.
In order to solve the technical problems, the invention provides the following technical scheme:
a Polar code decoding method comprises the following steps:
acquiring a channel receiving sequence with decoding failure;
generating disturbance noise matched with the channel receiving sequence;
correcting likelihood information of the channel receiving sequence by using the disturbance noise;
and decoding the channel receiving sequence by using the corrected likelihood information.
Preferably, the modifying the likelihood information of the channel receiving sequence by using the disturbance noise includes:
acquiring the deviation degree of the channel receiving sequence relative to a decoding error correction region after the disturbance noise is added;
and correcting the likelihood information of the channel receiving sequence by using the deviation degree.
Preferably, the generating disturbance noise matched with the channel receiving sequence comprises:
and inputting the channel receiving sequence into a trained neural network, and taking the output of the neural network as the disturbance noise.
Preferably, the process of training the neural network comprises:
generating a database with known samples; wherein each of the known samples comprises a neural network input and a neural network output, the neural network output being a perturbation noise that enables the neural network input to be correctly decoded;
and training the neural network by using the database until the loss value corresponding to the neural network reaches a preset threshold value.
Preferably, the method further comprises the following steps:
in iteratively training the neural network, the database is updated.
Preferably, decoding the channel received sequence by using the modified likelihood information includes:
inputting the likelihood information into an SC decoder, and decoding the channel receiving sequence by using the decoder to obtain an information estimation value;
performing CRC on the information estimation value;
if the verification is successful, the information estimation value is used as a decoding result;
if the check fails, judging whether the iterative decoding times reach an iterative threshold value; if not, repeatedly executing the step of generating the disturbance noise matched with the channel receiving sequence; if so, stopping iterative decoding.
Preferably, before the acquiring the channel receiving sequence with decoding failure, the method includes:
decoding the channel receiving sequence by using an SC decoder to obtain an information estimation value;
performing CRC on the information estimation value;
if the verification is successful, determining the information estimation value as a decoding result;
and if the verification fails, determining that the decoding of the channel receiving sequence fails.
A Polar code decoding device comprises:
a decoding failure sequence acquisition module, configured to acquire a channel receiving sequence with decoding failure;
the disturbing noise generating module is used for generating disturbing noise matched with the channel receiving sequence;
a likelihood information modifying module, configured to modify likelihood information of the channel receiving sequence by using the disturbance noise;
and the decoding module is used for decoding the channel receiving sequence by using the corrected likelihood information.
A signal receiver, comprising:
a memory for storing a computer program;
and the processor is used for realizing the steps of the Polar code decoding method when executing the computer program.
A readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the Polar code decoding method.
By applying the method provided by the embodiment of the invention, a channel receiving sequence with decoding failure is obtained; generating disturbance noise matched with a channel receiving sequence; correcting likelihood information of a channel receiving sequence by using the disturbance noise; and decoding the channel receiving sequence by using the corrected likelihood information.
It is considered that, assuming that relatively few errors occur, the channel reception sequence (i.e., the reception signal) is likely to be close to a certain error correction region, and if the channel reception sequence falls within the error correction region, no decoding error occurs. Therefore, only the channel receiving sequence with decoding failure needs to be shifted to the error correction area, so that the decoding error can be avoided and the decoding success rate can be improved. In the method, the channel receiving sequence with the decoding failure can be shifted to an error correction area based on the stochastic resonance phenomenon by generating the disturbance noise corresponding to the channel receiving sequence with the decoding failure. That is, after the disturbance noise is obtained, the likelihood information of the corresponding channel receiving sequence can be corrected by using the disturbance noise, and then the channel receiving sequence is decoded by following the likelihood information. Therefore, in the method, even if the decoding fails, the decoding can be carried out by generating the corresponding disturbance noise, so that the decoding success rate can be improved.
Correspondingly, the embodiment of the invention also provides a Polar code decoding device, equipment and a readable storage medium corresponding to the Polar code decoding method, which have the technical effects and are not described herein again.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flowchart illustrating an implementation of a Polar code decoding method according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating a relationship between a channel-received codeword and an associated error correction region according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating an embodiment of a Polar code decoding method according to the present invention;
FIG. 4 is a diagram illustrating a database update according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of a Polar code decoding apparatus according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating a structure of a signal receiver according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of a signal receiver according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the disclosure, reference will now be made in detail to the embodiments of the disclosure as illustrated in the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a flowchart of a Polar code decoding method according to an embodiment of the present invention.
In order to make those skilled in the art better understand the Polar code decoding method provided by the embodiment of the present invention, the following explains the core idea and related principles of the method.
First, the presence of noise is undesirable in a communication system because the noise can corrupt the transmitted information so that it is not recoverable at the receiver. However, it has been found that adding corresponding disturbing noise to the received signal (i.e. the channel-received sequence, i.e. the sequence of code words received by the channel) is rather beneficial for the decoding process. Specifically, referring to fig. 2, fig. 2 shows the relationship between the received code words of the channel and the associated error correction regions. In fig. 2, different error correction regions a ═ of M valid codewords are defined (a ═1，a2，...，am). For and error correction area a2And if the channel received signal y falls within the error correction area, no decoding error occurs, and if the channel received signal y falls outside the error correction area, a decoding error occurs. As can be seen from fig. 2, the disturbance noise n can shift the channel reception signal y to the correct error correction region. Assuming that relatively few errors occur, the channel received signal y is likely to be close to its designated error correction region. The phenomenon of stochastic resonance, i.e. the enhancement of the signal by adding random independent noise, may explain this beneficial effect of disturbing noise.
The specific implementation process comprises the following steps:
s101, acquiring a channel receiving sequence with decoding failure.
The three elements of the communication system are: a source (i.e., originating device, transmitting end), a sink (receiving end device, receiving end), and a channel (transmission medium). In this embodiment, at the transmitting end, information bits u with length K are encoded into a binary Polar code X with length N by adding CRC redundancy and Polar code encoder, and the codeword X is mapped to a symbol vector S by BPSK modulation, and BPSK signal will be transmitted through a channel with additive noise.
Specifically, before executing step S101, the method includes:
step one, decoding a channel receiving sequence by using an SC decoder to obtain an information estimation value;
performing CRC on the information estimation value;
step three, if the verification is successful, determining the information estimation value as a decoding result;
and step four, if the check fails, determining that the decoding of the channel receiving sequence fails.
For convenience of description, the above four steps will be described in combination.
At the receiving end, the received information y may be represented by equation (1):
y＝s+n (1)
referring to fig. 3, the SC decoder is first used to decode the channel received sequence y, and the information estimation value is obtained by the decision of equation (2):
wherein h isiFor the decision function, the following formula (3) is defined:
in the decoding process, the ratio of transition probabilities is defined as Likelihood Ratio (LR), i.e. likelihood information, i.e. LR of a single bit:
the iterative calculation formula of LR is shown in equations (5) and (6):
when decoding the ith bit, if i ∈ AcThen directly estimate the value
Assigned a value of ui(ii) a If i ∈ A, then the previous estimate needs to be based on
Calculating LR of current decoding bit according to LR iterative calculation formula, and then performing decision according to formula (7) to obtain information estimation value
If the information estimation value passes through CRC check, the decoding is successful, and a decoding result is output; if the information estimation value fails CRC check, decoding failure is indicated.
After the decoding fails, the channel receiving sequence may be used as a processing object, that is, the channel receiving sequence corresponding to step S101 is obtained.
And S102, generating disturbance noise matched with the channel receiving sequence.
The disturbing noise interferes with the channel reception sequence, and may cause the channel reception sequence to move to an error correction region. In order to enable the channel receiving sequence to be decoded effectively, when the disturbance noise is generated, the generation may be performed based on the channel receiving sequence, for example, the disturbance noise matched with the characteristics of the channel receiving sequence is generated.
Preferably, in order to further improve the effect of the disturbance noise on decoding, a neural network (CNN) learning capability can be further utilized to obtain the disturbance noise capable of shifting the channel receiving sequence to a correct error correction region through training. That is, the disturbance noise matched with the channel receiving sequence is generated, specifically, the channel receiving sequence is input to a trained neural network, and the output of the neural network is used as the disturbance noise. That is, the channel reception sequence is characterized by CNN, and disturbance noise is generated.
The training process of the neural network comprises the following steps:
stage one: generating a database with known samples; wherein each known sample comprises a neural network input and a neural network output, and the neural network output is disturbance noise which enables the neural network input to be correctly decoded;
and a second stage: and training the neural network by using the database until the loss value corresponding to the neural network reaches a preset threshold value.
Preferably, in order to further optimize the neural network, the database is also updated in the iterative training of the neural network.
The neural network utilizes a layered model structure similar to the human brain to extract features from input data from a bottom layer to a high layer step by step, and a mapping relation is established between a bottom layer signal and high-layer semantics.
The training quality is affected by some hyper-parameters, such as the size of the training set, small batch size, etc., but in this embodiment, it is possible to select only one set of valid hyper-parameters without paying attention to the optimization of these hyper-parameters. For example, the parameters shown in the following table may be used:
| CNN structure | {4；9,3,3,5；64,32,16,1} |
| Mini-batch size | 500 |
| Amount of training data | 700000 |
| Amount of test data | 200000 |
| SNR for generating data | {0,0.5,1,1.5,2,2.5,3,3.5,4}dB |
| Initialization method | Xavier |
| Optimization method | Adam |
The training neural network is divided into two stages to train the neural network, and the whole training process is finished off line, so that the complexity of decoding cannot be increased. A normality test of the loss function is used to measure the probability of the received value in decoding the error correction space, and the calculation of the likelihood information is facilitated later. The loss function is defined as:
wherein skewness S and kurtosis K are defined as follows:
wherein n isiRepresenting the ith element, n 'in the true noise vector'iIs the ith element of the CNN output vector,
the mean of the samples is shown. Through simulation experiments, the loss function can provide a very ideal training output. The smaller the loss value calculated by the loss function, the more similar the actual neural network output is to the expected network output, which means better training of the network.
In the first training phase, a database is generated in which the noise of each sample is known. Specifically, each sample in the database includes y (as an input to the neural network), i.e., the channel receive sequence, and perturbation noise (as an expected output of the neural network) that enables it to be correctly decoded.
Firstly, a large amount of simulation can be carried out, Polar codes are used for encoding random information bits, then, noise BPSK symbols are decoded by SC, if CRC (cyclic redundancy check) cannot pass, decoding failure is indicated, and disturbance noise needs to be added. In this case, y and the perturbation noise (taking the negative of the added random noise) that can be returned to the decoding error correction region can be stored as a known sample.
The neural network is then trained using the training data to minimize the loss function. To train the network correctly, the known samples in the database can be divided into two groups: training set and validation set. The training set is used to train the network, while the validation set is used to avoid overfitting.
Considering the situation that the decoding still fails after the disturbance noise is added, the loop iteration is carried out, the disturbance noise is repeatedly added, the first stage is untrained, and the neural network is further trained to output the correct disturbance noise on the basis of the last disturbance noise. To address this problem, CNN training continues in the second phase, as shown in fig. 4. In each iteration, a new database is generated based on the CNN. The neural network is then trained using reinforcement learning based on the new database.
The input of the CNN is a channel receiving value, that is, a channel receiving sequence, which is a one-dimensional vector, a rapid start Sequential (Sequential) model in a keras library is available, the number of convolution kernels of the first layer of the network is 64, the size of the convolution kernels is 9, and the convolution mode is selected to be same, so that the output of the first layer is 64 columns and n rows. Similarly, the number of convolution kernels in the second layer is 32, the size of the convolution kernels is 3, the number of convolution kernels in the third layer is 16, the size of the convolution kernels is 3, the number of convolution kernels in the fourth layer is 1, and the size of the convolution kernels is 15. The activation function of the first three layers is RELU, and the activation function of the fourth layer is linear, so that after the four-layer network is passed, the output result of CNN is ideal disturbance noise.
And S103, correcting likelihood information of the channel receiving sequence by using the disturbance noise.
Specifically, the process for correcting the likelihood information includes:
acquiring the deviation degree of a channel receiving sequence relative to a decoding error correction area after disturbance noise is added;
and step two, correcting the likelihood information of the channel receiving sequence by using the deviation degree.
Specifically, the LR is then modified according to the added disturbance noise,
wherein p is0,iAnd p1,iThe probability that the ith element of y' is-1 or 1. The calculation formula is as follows,
here, prob is the degree of deviation of the received value from the decoding error correction space after adding the disturbance noise, that is, the degree of deviation. The degree of deviation may be generated and stored in advance when the neural network is trained,
wherein the content of the first and second substances,
represent
M is-10 and q is 0.01.
And S104, decoding the channel receiving sequence by using the corrected likelihood information.
Specifically, the updated likelihood information is input to the SC decoder to decode the channel received sequence.
Decoding the channel received sequence using the modified likelihood information, comprising:
inputting likelihood information into an SC decoder, and decoding a channel receiving sequence by using the decoder to obtain an information estimation value;
performing CRC on the information estimation value;
step three, if the verification is successful, taking the information estimation value as a decoding result;
if the check fails, judging whether the iterative decoding times reach an iterative threshold value; if not, repeatedly executing the step of generating the disturbance noise matched with the channel receiving sequence; if so, stopping iterative decoding.
Inputting the updated likelihood information into SC decoder to obtain information estimation value
Judging whether the decoding is successful or not, if the decoding is failed, repeatedly adding disturbance noise until the decoding is successful or the maximum iteration number is reached. The more the iteration times, the better the decoding performance, and the decoding time delay is increased. The basis is to ensure the lowest decoding delay while obtaining better decoding performance. For example, the maximum number of iterations maySet to 10, 20, etc.
In particular, the process of decoding based on the likelihood information after correction may refer to the process of decoding using the likelihood information before correction.
In order to understand the relationship between decoding by using likelihood information before correction and decoding by using likelihood information after correction, please refer to fig. 3, and the decoding process will be described in detail from the received signal.
Step 1, carrying out serial offset decoding on the channel receiving sequence to obtain an information estimation value.
And 2, performing CRC (cyclic redundancy check) on the information estimation value, if the information estimation value passes the CRC, proving that the decoding is successful, and outputting decoding information (namely outputting the information estimation value as a decoding result), otherwise, performing the step 3.
Step 3, judging whether the maximum iteration times is reached, if so, terminating iterative decoding; otherwise, inputting the decoding failure sequence into the trained neural network, and then outputting the disturbance noise which can enable the decoding failure sequence to move to the decoding error correction area by the CNN and storing the disturbance noise.
And 4, correcting and storing the LR value of the decoding failure sequence according to the added disturbance noise.
And 5, carrying out SC decoding according to the updated LR to obtain an information estimation value, and turning to the step 2 to carry out CRC.
By applying the method provided by the embodiment of the invention, the channel receiving sequence with decoding failure is obtained; generating disturbance noise matched with a channel receiving sequence; correcting likelihood information of a channel receiving sequence by using the disturbance noise; and decoding the channel receiving sequence by using the corrected likelihood information.
It is considered that, assuming that relatively few errors occur, the channel reception sequence (i.e., the received signal) is likely to be close to a certain error correction region, whereas if the channel reception sequence falls within the error correction region, no decoding error occurs. Therefore, only the channel receiving sequence with decoding failure needs to be shifted to the error correction area, so that the decoding error can be avoided and the decoding success rate can be improved. In the method, the channel receiving sequence with failed decoding can be shifted to an error correction region based on the stochastic resonance phenomenon by generating the disturbance noise corresponding to the channel receiving sequence with failed decoding. After the disturbance noise is obtained, the likelihood information of the corresponding channel receiving sequence can be corrected by using the disturbance noise, and then the channel receiving sequence is decoded by following the likelihood information. Therefore, in the method, even if the decoding fails, the decoding can be carried out by generating the corresponding disturbance noise improvement, so that the decoding success rate can be improved.
Corresponding to the above method embodiments, the embodiment of the present invention further provides a Polar code decoding apparatus, and the Polar code decoding apparatus described below and the Polar code decoding method described above may be referred to correspondingly.
Referring to fig. 5, the apparatus includes the following modules:
a decoding failure sequence acquisition module 101, configured to acquire a channel receiving sequence with decoding failure;
a disturbing noise generating module 102, configured to generate a disturbing noise matched with a channel receiving sequence;
a likelihood information modifying module 103, configured to modify likelihood information of the channel receiving sequence by using the disturbance noise;
and a decoding module 104, configured to decode the channel received sequence by using the corrected likelihood information.
By applying the device provided by the embodiment of the invention, the channel receiving sequence with decoding failure is obtained; generating disturbance noise matched with a channel receiving sequence; correcting likelihood information of a channel receiving sequence by using the disturbance noise; and decoding the channel receiving sequence by using the corrected likelihood information.
It is considered that, assuming that relatively few errors occur, the channel reception sequence (i.e., the reception signal) is likely to be close to a certain error correction region, and if the channel reception sequence falls within the error correction region, no decoding error occurs. Therefore, only the channel receiving sequence with decoding failure needs to be shifted to the error correction area, so that the decoding error can be avoided and the decoding success rate can be improved. In the present apparatus, by generating a disturbance noise corresponding to a channel received sequence whose decoding has failed, the channel received sequence whose decoding has failed can be shifted to an error correction region based on a stochastic resonance phenomenon. After the disturbance noise is obtained, the likelihood information of the corresponding channel receiving sequence can be corrected by using the disturbance noise, and then the channel receiving sequence is decoded by following the likelihood information. Therefore, even if the decoding fails, the device can generate corresponding disturbance noise to improve decoding, and can improve the success rate of decoding.
In a specific embodiment of the present invention, the likelihood information modifying module 103 is specifically configured to obtain a deviation degree of the channel receiving sequence after adding the disturbance noise with respect to a decoding error correction region; and correcting the likelihood information of the channel receiving sequence by using the deviation degree.
In an embodiment of the present invention, the disturbance noise generation module 102 is specifically configured to input the channel receiving sequence to a trained neural network, and use an output of the neural network as the disturbance noise.
In one embodiment of the invention, the neural network training module is used for generating a database with known samples; wherein each known sample comprises a neural network input and a neural network output, the neural network output being a perturbation noise that enables the neural network input to be correctly decoded; and training the neural network by using the database until the loss value corresponding to the neural network reaches a preset threshold value.
In an embodiment of the present invention, the neural network training optimization module is specifically configured to update the database in the iterative training of the neural network.
In an embodiment of the present invention, the decoding module 104 is specifically configured to input the likelihood information into an SC decoder, and decode the channel receiving sequence by using the SC decoder to obtain an information estimation value; performing CRC on the information estimation value; if the verification is successful, the information estimation value is used as a decoding result; if the check fails, judging whether the iterative decoding times reach an iterative threshold value; if not, repeatedly executing the step of generating the disturbance noise matched with the channel receiving sequence; if so, stopping iterative decoding.
In a specific embodiment of the present invention, the decoding module 104 is configured to decode the channel receiving sequence by using an SC decoder before acquiring the channel receiving sequence with decoding failure, so as to obtain an information estimation value; performing CRC on the information estimation value; if the verification is successful, determining the information estimation value as a decoding result; and if the check fails, determining that the decoding of the channel receiving sequence fails.
Corresponding to the above method embodiments, the embodiment of the present invention further provides a signal receiver, and a signal receiver described below and a Polar code decoding method described above may be referred to with each other.
Referring to fig. 6, the signal receiver includes:
a memory D1 for storing computer programs;
and a processor D2, configured to implement the steps of the Polar code decoding method in the foregoing method embodiments when executing the computer program.
Specifically, referring to fig. 7, a specific structural diagram of a signal receiver provided in this embodiment is a schematic diagram of a signal receiver, which may generate relatively large differences due to different configurations or performances, and may include one or more processors (CPUs) 322 (e.g., one or more processors), a memory 332, and one or more storage media 330 (e.g., one or more mass storage devices) storing an application 342 or data 344. Memory 332 and storage media 330 may be, among other things, transient storage or persistent storage. The program stored on the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions operating on a data processing device. Still further, the central processor 322 may be configured to communicate with the storage medium 330 to execute a series of instruction operations in the storage medium 330 on the signal receiver 301.
The signal receiver 301 may also include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input-output interfaces 358, and/or one or more operating systems 341. Such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.
The steps in the Polar code decoding method described above may be implemented by the structure of a signal receiver.
Corresponding to the above method embodiment, an embodiment of the present invention further provides a readable storage medium, and a readable storage medium described below and a Polar code decoding method described above may be referred to in correspondence.
A readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the Polar code decoding method of the foregoing method embodiment.
The readable storage medium may be a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, an optical disk, or other various readable storage media capable of storing program codes.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
Claims (9)
1. A Polar code decoding method is characterized by comprising the following steps:
acquiring a channel receiving sequence with decoding failure;
generating disturbance noise matched with the channel receiving sequence;
correcting likelihood information of the channel receiving sequence by using the disturbance noise;
decoding the channel receiving sequence by using the corrected likelihood information;
the generating of the disturbance noise matched with the channel receiving sequence comprises:
inputting the channel receiving sequence into a trained neural network, and taking the output of the neural network as the disturbance noise.
2. The decoding method of claim 1, wherein the step of correcting the likelihood information of the channel receiving sequence by using the interference noise comprises:
acquiring the deviation degree of the channel receiving sequence relative to a decoding error correction region after the disturbance noise is added;
and correcting the likelihood information of the channel receiving sequence by using the deviation degree.
3. The Polar code decoding method according to claim 1, wherein the process of training the neural network comprises:
generating a database with known samples; wherein each of the known samples comprises a neural network input and a neural network output, the neural network output being a perturbation noise that enables the neural network input to be correctly decoded;
and training the neural network by using the database until the loss value corresponding to the neural network reaches a preset threshold value.
4. The decoding method of Polar code according to claim 3, further comprising:
in iteratively training the neural network, the database is updated.
5. The decoding method of Polar code according to claim 1, wherein decoding the received channel sequence using the revised likelihood information comprises:
inputting the likelihood information into an SC decoder, and decoding the channel receiving sequence by using the decoder to obtain an information estimation value;
performing CRC on the information estimation value;
if the verification is successful, the information estimation value is used as a decoding result;
if the check fails, judging whether the iterative decoding times reach an iterative threshold value; if not, repeatedly executing the step of generating the disturbance noise matched with the channel receiving sequence; if so, stopping iterative decoding.
6. The decoding method of Polar code according to claim 1, wherein before said obtaining the channel receiving sequence with decoding failure, it includes:
decoding the channel receiving sequence by using an SC decoder to obtain an information estimation value;
performing CRC on the information estimation value;
if the verification is successful, determining the information estimation value as a decoding result;
and if the verification fails, determining that the decoding of the channel receiving sequence fails.
7. A Polar code decoding device, comprising:
a decoding failure sequence acquisition module for acquiring a channel receiving sequence with decoding failure;
the disturbing noise generating module is used for generating disturbing noise matched with the channel receiving sequence;
a likelihood information modifying module for modifying likelihood information of the channel receiving sequence by using the disturbance noise;
a decoding module, configured to decode the channel receiving sequence by using the corrected likelihood information;
the disturbance noise generation module is specifically configured to: inputting the channel receiving sequence into a trained neural network, and taking the output of the neural network as the disturbance noise.
8. A signal receiver, comprising:
a memory for storing a computer program;
processor for implementing the steps of the Polar code decoding method according to any one of claims 1 to 6 when executing the computer program.
9. Readable storage medium, wherein a computer program is stored on the readable storage medium, and when being executed by a processor, the computer program implements the steps of the Polar code decoding method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010229739.5A CN111200441B (en) | 2020-03-27 | 2020-03-27 | Polar code decoding method, device, equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010229739.5A CN111200441B (en) | 2020-03-27 | 2020-03-27 | Polar code decoding method, device, equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111200441A CN111200441A (en) | 2020-05-26 |
| CN111200441B true CN111200441B (en) | 2022-07-15 |
Family
ID=70746979
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010229739.5A Active CN111200441B (en) | 2020-03-27 | 2020-03-27 | Polar code decoding method, device, equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111200441B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114389620A (en) * | 2020-10-20 | 2022-04-22 | 华为技术有限公司 | Decoding method and device |
| US20240080045A1 (en) * | 2021-01-08 | 2024-03-07 | Nippon Telegraph And Telephone Corporation | Coding circuit, decoding circuit, coding method, decoding method and computer program |
| CN112929129B (en) * | 2021-02-09 | 2022-07-12 | 中国人民解放军战略支援部队信息工程大学 | Data processing system, method and equipment based on dynamic redundancy heterogeneous coding |
| CN113271110B (en) * | 2021-04-27 | 2022-12-06 | 北京邮电大学 | Semantic information source channel joint coding and decoding method and device for text |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107026656A (en) * | 2017-01-10 | 2017-08-08 | 南京邮电大学 | A kind of long Polar codes efficient coding method of short code in CRC auxiliary based on disturbance |
| CN109617655A (en) * | 2018-12-28 | 2019-04-12 | 中北大学 | A kind of polarization code wireless data safe transmission method based on random scrambling code auxiliary |
| CN110798228A (en) * | 2019-10-29 | 2020-02-14 | 南京宁麒智能计算芯片研究院有限公司 | Polarization code turning decoding method and system based on deep learning |
-
2020
- 2020-03-27 CN CN202010229739.5A patent/CN111200441B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107026656A (en) * | 2017-01-10 | 2017-08-08 | 南京邮电大学 | A kind of long Polar codes efficient coding method of short code in CRC auxiliary based on disturbance |
| CN109617655A (en) * | 2018-12-28 | 2019-04-12 | 中北大学 | A kind of polarization code wireless data safe transmission method based on random scrambling code auxiliary |
| CN110798228A (en) * | 2019-10-29 | 2020-02-14 | 南京宁麒智能计算芯片研究院有限公司 | Polarization code turning decoding method and system based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 基于PD-CNN的Polar码译码算法;徐鹏等;《信号处理》;20191031;第35卷(第10期);第1652-1659页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111200441A (en) | 2020-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111200441B (en) | Polar code decoding method, device, equipment and readable storage medium | |
| CN108462558B (en) | Method and device for decoding polarization code SCL and electronic equipment | |
| US20180357530A1 (en) | Deep learning decoding of error correcting codes | |
| CN102224679B (en) | Encoding apparatus, decoding apparatus, encoding method, decoding method, and communication system | |
| US11121728B2 (en) | Pre-coding and decoding polar codes using local feedback | |
| US10784992B2 (en) | Device and method for executing encoding | |
| EP3539237B1 (en) | Error detection in communication systems using polar coded data transmission | |
| US10797728B1 (en) | Systems and methods for diversity bit-flipping decoding of low-density parity-check codes | |
| US20200007170A1 (en) | System and method for early termination of decoding in a multi user equipment environment | |
| US11847019B2 (en) | Polar code construction method and apparatus | |
| WO2020108586A1 (en) | Polar code decoding method and apparatus, multi-stage decoder, and storage medium | |
| US20230179229A1 (en) | List decoding of polarization-adjusted convolutional codes | |
| CN116530023A (en) | Serial concatenated code with outer block code and inner polarization-adjusted convolutional code | |
| WO2019030531A1 (en) | Apparatus and method for detecting mutually interfering information streams | |
| CN112953569B (en) | Decoding method and device, storage medium, electronic equipment and decoder | |
| WO2018157263A1 (en) | Generalized polar codes | |
| CN112491422A (en) | Gauss optimization-based bit flipping serial elimination list decoding method and system | |
| CN111835363B (en) | LDPC code decoding method based on alternate direction multiplier method | |
| CN110708077B (en) | LDPC code large number logic decoding method, device and decoder | |
| CN112615629B (en) | Decoding method, system, medium, equipment and application of multi-element LDPC code | |
| CN112511172B (en) | Decoding method, device, equipment and storage medium | |
| CN114844511A (en) | Polar code decoding method and device, storage medium and electronic equipment | |
| CN110830166B (en) | Joint detection decoding method and device, computer equipment and storage medium | |
| Kobozeva et al. | Investigation of signal-code structures based on 3D error-locating codes | |
| JP2010535459A (en) | Coordinate ascent method for linear programming decoding. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |