KR20230156883A - Apparatus and method for decoding of polar codes using future constraints in communication system or broadcasting system - Google Patents

Abstract

In embodiments of the present disclosure, a device performed by a device in a communication system or broadcasting system includes: memory; at least one transceiver; And it may include at least one processor. The at least one processor may be configured to receive a signal including bits encoded based on polar codes. The at least one processor may be configured to identify polar code configurations for at least one information bit and at least one frozen bit of the signal. The at least one processor may be configured to decode the signal based on the configuration of the polar code. Decoding of an information bit among the at least one information bit may be performed based on at least one value of at least one constraint bit. The at least one constraint bit may include at least one of a parity bit or a freeze bit corresponding to a later order than the decoding order of the information bits.

Description

Apparatus and method for decoding polar codes using future constraints in a communication system or broadcasting system {APPARATUS AND METHOD FOR DECODING OF POLAR CODES USING FUTURE CONSTRAINTS IN COMMUNICATION SYSTEM OR BROADCASTING SYSTEM}

This disclosure generally relates to an apparatus and method for decoding a signal transmitted using a polar code in a communication system or broadcasting system. Specifically, the present disclosure relates to using at least one future constraint in an apparatus and method for decoding a transmitted signal using polar codes.

In general, when data is transmitted and received between a transmitter and a receiver in a communication and broadcasting system, data errors may occur due to noise present in the communication channel. As coding methods designed to enable the receiver to process errors generated by communication channels, there are error detection codes (error detection codes) and error correcting codes (ECC). The error detection code method is a technique that allows the receiver to check whether the received data contains errors, and the error correction code method is a technique that allows the receiver to self-correct errors included in the received data. Here, the error correction coding method is often called channel coding or forward error correction (FEC).

There are various methods of error correction coding techniques. Representative error correction codes include convolutional code, turbo code, low-density parity-check code (LDPC code), and polar code method. In particular, turbo codes, low-density parity check codes, and polar codes are excellent channel codes that have or achieve performance close to theoretical channel capacity and are used in various communication and broadcasting systems today.

References described in this disclosure are as follows.

[1] E. Arikan, “Channel Polarization: a method for constructing capacity-achieving codes for symmetric binary-input memoryless channels,” IEEE Trans. Information Theory , vol. 55, no. 7, pp. 3051-3073, July 2009.

[2] 3GPP, NR multiplexing and channel coding (Release 15), TSG RAN TS38.212 v15.0.1, Feb. 2017.

[3] A. Balatsoukas-Stimming, M. B. Parizi, and A. Burg, “LLR-based successive cancellation list decoding of polar codes,” IEEE Trans. Sig. Process., vol. 63, no. 19, pp. 5165 - 5179, June 2015.

Among the error correction code techniques, polar code is a channel code that achieves point-to-point channel capacity in a simple and effective manner by using a phenomenon called channel polarization [1]. In the process of transmitting multiple bits through an independent bit channel, when encoding using a structured generator matrix and sequential cancellation (SC) decoding are used, the channel for each bit is virtual. It is transformed into a polarized synthesized channel. In this process, some synthetic channels become excellent channels with a maximum channel capacity close to 1, while other synthetic channels become poor channels with a minimum channel capacity close to 0. The total sum of the channel capacities of the composite channels remains the same before and after the change. And as the code length increases, channel polarization is maximized, so a good channel has a channel capacity of 1, and a poor channel has a channel capacity of 0. Therefore, theoretically, the transmitter can easily and effectively achieve channel capacity for a given channel by transmitting information bits to be transmitted to good channels and allocating frozen bits to bad channels.

The SC decoding method for the polar code is modified to a similar maximum likelihood (near-ML or ML-like) decoding method such as SCL (SC-list) decoding, SCS (SC-stack) decoding, and SC-flip (SCF) decoding. Easy to expand. This improved decoding algorithm achieves better error correction performance. For this reason, 3GPP NR, a 5G communication standard, uses polar codes when transmitting short-length control information [2].

This disclosure considers communication and broadcasting systems that use polar codes. The basic decoding method for polar codes is sequential cancellation (SC), which is characterized by sequentially decoding the encoded input bit sequence one bit at a time based on the previously estimated values of the bits. The SC decoding is being modified and expanded with improved algorithms such as SCL (SC-list) decoding, SCS (SC-stack) decoding, and SCF (SC-flip) decoding. These algorithms also basically sequentially process bits. It is based on a sequential removal method of decoding.

The present disclosure relates to a method and device for improving characteristics and performance by realistically overcoming aspects that have not been considered in the conventional SC-based sequential decoding of polar codes. The SC decoding proposed in [1] is established by assuming that in the step of decoding a specific bit, the bits to be decoded thereafter are unconstrained and can have a value of 0 or 1. However, when forming a polar code, there may be at least one constrained bit such as a frozen bit (or fixed bit) or a parity bit among the bits to be decoded. Therefore, in the step of decoding a specific bit, at least one restricted bit may be included among the bits to be decoded thereafter. SC decoding is not strictly an optimal decoding method because it does not consider the restricted bits that exist after the code is constructed. This disclosure has discovered this aspect, and presents methods and embodiments of improving the SC decoding method and device by considering constraints that exist after the decoding of a specific bit, or future constraints for short, in a realistic way. .

According to embodiments of the present disclosure, a method performed by a device in a communication system or broadcasting system may include receiving a signal including bits encoded based on polar codes. The method may include identifying a polar code configuration for at least one information bit and at least one frozen bit of the signal. The method may include performing decoding of the signal based on the configuration of the polar code. Decoding of an information bit among the at least one information bit may be performed based on at least one value of at least one constraint bit. The at least one constraint bit may include at least one of a parity bit or a freeze bit corresponding to a later order than the information bit.

In embodiments of the present disclosure, a device performed by a device in a communication system or broadcasting system includes: memory; at least one transceiver; And it may include at least one processor. The at least one processor may be configured to receive a signal including bits encoded based on polar codes. The at least one processor may be configured to identify polar code configurations for at least one information bit and at least one frozen bit of the signal. The at least one processor may be configured to decode the signal based on the configuration of the polar code. Decoding of an information bit among the at least one information bit may be performed based on at least one value of at least one constraint bit. The at least one constraint bit may include at least one of a parity bit or a freeze bit corresponding to a later order than the information bit.

The apparatus and method according to embodiments of the present disclosure can improve error correction performance by performing decoding based on at least one future constraint.

The effects that can be obtained from the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below. will be.

1 illustrates a communication system or broadcasting system according to embodiments of the present disclosure.
Figure 2 shows an example of the configuration of a receiving end device in a communication system or broadcasting system according to embodiments of the present disclosure.
Figure 3 shows an example of a graph corresponding to a polar sign in a communication system or broadcasting system according to embodiments of the present disclosure.
FIG. 4 illustrates an element process for performing log-likelihood ratio (LLR) calculation and sequential removal in a polar code decoding method and device for a communication system or broadcasting system according to an embodiment of the present disclosure.
FIG. 5 illustrates an example of a binary decision tree for sequentially decoding and estimating bits in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure.
FIG. 6 illustrates an example of a binary decision tree considered in a polar code decoding method and apparatus of a communication system or broadcasting system according to embodiments of the present disclosure.
Figure 7 shows the performance of a decoding technique according to embodiments of the present disclosure.
FIG. 8 shows an example of a binary decision tree used in a polar code decoding method and apparatus of a communication system or broadcasting system according to embodiments of the present disclosure.
FIG. 9 illustrates an example of a graph modified in consideration of future constraints in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure.
FIG. 10 illustrates examples of graphs considering future constraints configured when decoding each bit in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure.
FIG. 11 illustrates examples of graphs considering future constraints configured when decoding each bit in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure.
Figure 12 shows the operational flow of a device for decoding based on future constraints according to embodiments of the present disclosure.
FIG. 13 shows an example of a concatenated polar coding system using outer coding according to embodiments of the present disclosure.
Figure 14 shows an example of an equivalent model of a concatenated polar code system using external coding according to embodiments of the present disclosure.
Figure 15 shows an example of an intermediate codeword vector in embodiments of the present disclosure.

Hereinafter, embodiments of the present disclosure are described in detail along with the attached drawings.

In describing the embodiments, descriptions of technical content that is well known in the technical field to which the present invention belongs and that are not directly related to the present invention will be omitted. This is to convey the gist of the present invention more clearly without obscuring it by omitting unnecessary explanation.

For the same reason, some components are emphasized, omitted, or schematically shown in the accompanying drawings. Additionally, the size of each component does not entirely reflect its actual size. In each drawing, identical or corresponding components are assigned the same reference number.

The advantages and features of the present invention and methods for achieving them will become clear by referring to the embodiments described in detail below along with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and the present embodiments are merely intended to make the disclosure of the present invention complete, and are not limited to the embodiments disclosed below, and are provided by those skilled in the art in the technical field to which the present invention pertains. It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. The same reference numerals may refer to the same elements throughout the specification.

At this time, it will be understood that each block of the processing flow diagram diagrams and combinations of the flow diagram diagrams can be performed by computer program instructions. Since these computer program instructions can be mounted on a processor of a general-purpose computer, special-purpose computer, or other programmable data processing equipment, the instructions performed through the processor of the computer or other programmable data processing equipment are described in the flowchart block(s). It creates the means to perform functions. These computer program instructions may also be stored in computer-usable or computer-readable memory that can be directed to a computer or other programmable data processing equipment to implement a function in a particular manner, so that the computer-usable or computer-readable memory It is also possible to produce manufactured items containing instruction means that perform the functions described in the flowchart block(s). Computer program instructions can also be mounted on a computer or other programmable data processing equipment, so that a series of operational steps are performed on the computer or other programmable data processing equipment to create a process that is executed by the computer, thereby generating a process that is executed by the computer or other programmable data processing equipment. Instructions that perform processing equipment may also provide steps for executing the functions described in the flow diagram block(s).

Additionally, each block may represent a module, segment, or portion of code that includes one or more executable instructions for executing specified logical function(s). Additionally, it should be noted that in some alternative execution examples it is possible for the functions mentioned in the blocks to occur out of order. For example, it is possible for two blocks shown in succession to be performed substantially at the same time, or it is possible for the blocks to be performed in reverse order depending on the corresponding function.

At this time, the term '~unit' used in this embodiment refers to software or hardware components such as FPGA or ASIC, and the '~unit' performs certain roles. However, '~part' is not limited to software or hardware. The '~ part' may be configured to reside in an addressable storage medium and may be configured to reproduce on one or more processors. Therefore, as an example, '~ part' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. The functions provided within the components and 'parts' may be combined into a smaller number of components and 'parts' or may be further separated into additional components and 'parts'. In addition, the components and 'parts' may be implemented to regenerate one or more CPUs within the device or secure multimedia card.

Hereinafter, various embodiments will be described in detail with reference to the attached drawings. At this time, it should be noted that like components in the attached drawings are indicated by the same symbols whenever possible. In addition, it should be noted that the drawings of the present invention attached below are provided to aid understanding of the present invention, and the present invention is not limited to the form or arrangement illustrated in the drawings of the present invention. Additionally, detailed descriptions of well-known functions and configurations that may obscure the gist of the present invention will be omitted. It should be noted that in the following description, only the parts necessary to understand the operation according to various embodiments of the present invention will be described, and descriptions of other parts will be omitted so as not to confuse the gist of the present invention.

Terms referring to signals (e.g., signal, information, message, signaling), terms referring to resources, and terms for operational states (e.g., step, operation, procedure) used in the following description. ), terms referring to data (e.g. packets, user streams, information, bits, symbols, codewords), terms referring to channels, and network entities. Terms referring to devices, terms referring to components of a device, etc. are exemplified for convenience of explanation. Accordingly, the present disclosure is not limited to the terms described below, and other terms having equivalent technical meaning may be used.

In the present disclosure, expressions of more or less may be used to determine whether a specific condition is satisfied or fulfilled, but this is only a description for expressing an example and does not exclude descriptions of more or less. It's not. Conditions written as ‘more than’ can be replaced with ‘more than’, conditions written as ‘less than’ can be replaced with ‘less than’, and conditions written as ‘more than and less than’ can be replaced with ‘greater than and less than’.

The present disclosure describes some embodiments using terms used in some communication standards (e.g., 3rd Generation Partnership Project (3GPP)), but this is only an example for explanation. Various embodiments of the present disclosure may be used in other communication systems Also, it can be easily modified and applied.

1 shows a communication system and a broadcasting system according to embodiments of the present disclosure.

Referring to FIG. 1, a transmitting end 110 and a receiving end 120 are illustrated as parts of devices or nodes that use a wireless channel in a wireless communication system. Figure 1 shows one transmitting end 110 and one receiving end 120, but a communication system or a broadcasting system may include a plurality of transmitting ends or a plurality of receiving ends. Additionally, for convenience of explanation, in this disclosure, the transmitting end 110 and the receiving end 120 are described as separate entities, but the functions of the transmitting end 110 and the receiving end 120 may be interchanged. For example, in the case of the uplink of a cellular communication system, the transmitting end 110 may be a terminal and the receiving end 120 may be a base station. In the case of downlink, the transmitting end 110 may be a base station and the receiving end 120 may be a terminal.

A base station is a network infrastructure that provides wireless access to terminals. A base station has coverage defined based on the distance over which signals can be transmitted. In addition to the base station, the base station includes MMU (massive MIMO (multiple input multiple output) unit), 'access point (AP)', 'eNodeB (eNB)', and '5G node (5th generation node). ', '5G NodeB (NB)', 'wireless point', 'transmission/reception point (TRP)', 'access unit', 'distributed unit' Referred to as 'unit, DU)', 'transmission/reception point (TRP)', 'radio unit (RU), remote radio head (RRH)' or other terms with equivalent technical meaning. It can be. The base station can transmit a downlink signal or receive an uplink signal.

A terminal is a device used by a user and communicates with a base station through a wireless channel. In some cases, the terminal may be operated without user involvement. In other words, the terminal is a device that performs machine type communication (MTC) and may not be carried by the user. In addition to the terminal, the terminal includes 'user equipment (UE)', 'mobile station', 'subscriber station', 'customer premises equipment (CPE)', and 'remote'. ‘Remote terminal’, ‘wireless terminal’, ‘electronic device’, or ‘vehicle terminal’, ‘user device’ or equivalent technical meaning It may be referred to by other terms.

Although not shown in FIG. 1, in addition to communication between the base station and the terminal, the terminal may also perform direct communication with other terminals. For example, terminal vehicle communication different from the terminal shown in FIG. 1 may be supported. In the case of vehicle communication, in the LTE system, standardization work on V2X technology based on the device-to-device (D2D) communication structure was completed in 3GPP Release 14 and Release 15, and currently NR V2X in 5G NR Release 16. Standardization work on technology was underway.

Depending on the links formed between communication nodes, the transmitting end and the receiving end can be defined in various ways. According to one embodiment, the transmitting end 110 may be a base station, and the receiving end 120 may be a terminal. Additionally, according to another embodiment, the receiving end 120 may be a base station and the transmitting end 110 may be a terminal. Additionally, according to another embodiment, both the transmitting end and the receiving end may be terminals. Hereinafter, the present disclosure describes the subject transmitting the signal as the transmitting end and the subject receiving the signal as the receiving end, but this is only a functional expression to explain the signal processing process and is not interpreted as limiting a specific embodiment.

In embodiments, the transmitting end 110 may generate a codeword by encoding information bits based on polar codes, and the receiving end 120 may decode the signal of the received codeword based on the polar codes. there is. Subchannel allocation for input bits may be performed. Each input bit can be interpreted as if it passes through a split channel (sub-channel), a virtual channel of different quality, due to channel polarization. At this time, each subchannel is also called a synthetic channel. After allocating subchannels, the transmitter 110 may perform encoding based on polar codes using a generation matrix. The receiving end 120 may perform decoding based on polar codes through sequential decoding (SC).

In embodiments, the transmitter 110 may generate a codeword by encoding information bits based on an LDPC code. The receiving end 120 may decode the received codeword signal based on the LDPC code. For example, the receiving end 120 may use the LDPC decoding method according to the present invention and may perform a syndrome check to determine whether the decoding result is normal. The transmitter 110 may perform LDPC encoding using a parity check matrix. The receiving end 120 may perform LDPC decoding using a parity check matrix. For example, the parity check matrix may include a parity check matrix defined in the 3GPP NR standard.

Figure 2 shows an example of the configuration of a receiving end device in a communication system or broadcasting system according to embodiments of the present disclosure. That is, the configuration illustrated in FIG. 2 can be understood as the configuration of the receiving end 120. Terms such as '... unit' and '... unit' used hereinafter refer to a unit that processes at least one function or operation, which can be implemented through hardware, software, or a combination of hardware and software. there is.

Referring to FIG. 2, the device may include a communication unit 210, a storage unit 220, and a control unit 230.

The communication unit 210 may perform functions for transmitting and receiving signals through a wireless channel. For example, the communication unit 210 may perform a conversion function between a baseband signal and a bit string according to the physical layer standard of the system. For example, when transmitting data, the communication unit 210 may generate complex symbols by encoding and modulating the transmission bit string. Additionally, when receiving data, the communication unit 210 can restore the received bit stream by demodulating and decoding the baseband signal. Additionally, the communication unit 210 may up-convert the baseband signal into a radio frequency (RF) band signal and transmit it through an antenna, and down-convert the RF band signal received through the antenna into a baseband signal.

To this end, the communication unit 210 may include a transmission filter, a reception filter, an amplifier, a mixer, an oscillator, a digital to analog convertor (DAC), an analog to digital convertor (ADC), etc. Additionally, the communication unit 210 may include multiple transmission and reception paths. Furthermore, the communication unit 210 may include at least one antenna array composed of multiple antenna elements. In terms of hardware, the communication unit 210 may be composed of a digital unit and an analog unit, and the analog unit is divided into a number of sub-units according to operating power, operating frequency, etc. It can be configured. Additionally, the communication unit 210 may include a decoding unit to perform decryption according to various embodiments of the present disclosure.

The communication unit 210 transmits and receives signals as described above. Accordingly, the communication unit 210 may be referred to as a 'transmitter', 'receiver', or 'transceiver'. Additionally, in the following description, transmission and reception performed through a wireless channel are used to mean that the processing as described above is performed by the communication unit 210. Additionally, when the device in FIG. 2 is a base station, the communication unit 210 may further include a backhaul communication unit for communication with other network entities connected through a backhaul network.

The storage unit 220 may store data such as basic programs, application programs, and setting information for the operation of the receiving end 120. The storage unit 220 may be comprised of volatile memory, non-volatile memory, or a combination of volatile memory and non-volatile memory. Additionally, the storage unit 220 may provide stored data upon request from the control unit 230.

The control unit 230 can control overall operations of the device. For example, the control unit 230 may transmit and receive signals through the communication unit 210. Additionally, the control unit 230 can write or read data into the storage unit 220. To this end, the control unit 230 may include at least one processor or microprocessor, or may be part of a processor. According to various embodiments, the control unit 230 may control the device to perform operations according to various embodiments described later.

In the present disclosure, embodiments are described using terms used in some communication and broadcasting standards (e.g., 3 ^rd Generation Partnership Project (3GPP)), but this is only an example for explanation, and various embodiments of the present disclosure are used in other communications It can be easily modified and applied to the system as well.

Additionally, in the process of explaining the present invention in detail, generally accepted mathematical symbols are used. These mathematical symbols can be clearly understood by those skilled in the art to which this disclosure pertains. Typically, the following mathematical symbols are used in this disclosure.

● Calligraphic characters (e.g. ) is used to indicate a set.

● Throughout this disclosure, unless otherwise specified, it is assumed that the index of the first element of a set, sequence, or vector starts from 0 (zero-based numbering).

● symbol , , is used to refer to the set of natural numbers, the set of integers, and the set of real numbers, respectively.

● symbol represents a binary field.

● Non-negative integer About, starts from 0 in succession up to Represents a set of integers. in other words, am.

● Boldface lowercase letters (e.g. ) is used to indicate vectors, and boldface uppercase letters (e.g. ) is used to indicate a matrix. In the case of vectors, unless otherwise specified, they represent column vectors.

● Vector and matrix About and represents each transpose.

● Vector and two non-negative integers , About, is a vector of from the first element It represents a subvector composed of consecutive elements up to the th element. in other words, Is am.

The polar code has low coding/complexity performance that can be implemented with the error correction code proposed by E. Arikan, while also limiting data transmission performance in a binary discrete memoryless channel (B-DMC). It is the first error correction code proven to achieve in-channel capacity [1]. Decoding methods based on polar codes and sequential cancellation (SC) enable excellent error-correction performance when transmitting short-length codes compared to other channel codes. Because of these advantages, 3GPP New Radio (NR), the 5th generation (5G) mobile communication standard, uses polar codes to transmit short-length control information.

The operating principle of polar codes is channel polarization. Below, channel polarization is introduced based on what was described in [1]. Input alphabet (set of values the input can have) , output alphabet (the set of values the output can have) About B-DMC Channel is the channel input A channel output It is a function for transitioning to . Enter alphabet here is a binary field It's the same. Given channel input and channel output about Is When sending is the transition probability of receiving. length from transmitter In bit vector iid (independent and identically distributed) B-DMC Transmitted through and the receiver Consider receiving. length from here is a power of 2, any natural number About am. In other words, am. In this case the length The vector B-DMC of It can be defined as, Since each channel is independent, the length The transition probability of vector B-DMC is This happens.

Bit vectors transmitted in the communication systems and broadcasting systems considered in this disclosure is a codeword vector encoded with a polar sign. The encoder input bit vector is and the transmitter encodes this and obtains an encoded output bit vector. creates . The encoded output bit vector is also called a codeword vector, and in this disclosure, both names are used with the same meaning depending on the context. where the encoded input bit vector may include at least one parity bit as a result of being encoded by external encoding. The parity bit of the polar sign is characterized as being causally generated, specifically the parity bit. is the bit with an index smaller than this It is determined by a linear combination of these. Also encoded input bit vector may include at least one frozen bit (or frozen bit) to perform the target sign expurgation. The value of the freeze bit is pre-determined (i.e., not obtained as a result of the encoding and decoding process) and is usually 0.

The encoding process of the polar sign is the encoding input bit vector to generator matrix Multiply by the same length codeword vector of This is the process of creating a . Specifically, the codeword vector is generated by [Equation 1].

[Equation 1]

generation matrix was defined as in [Equation 2] in [1], where the first polar sign was proposed.

[Equation 2]

In [Equation 2] above, is called the polarization kernel. This is it. Above [Equation 2] superscript The calculation is It stands for Kronecker power. Kronecker power , It is calculated in an iterative manner. and silver size It is a bit-reversal permutation matrix. For example, a vector of length 8 go When multiplied by , the indices are bit-inverted vectors. is created. Recently, in various literature and systems, including the 3GPP NR system, Consider the generation matrix of [Equation 3] excluding .

[Equation 3]

Unless otherwise stated below, the generation matrix defined in [Equation 3] Assume. However, the contents described below and the contents of the present invention are the generation matrix defined in [Equation 2] The same can be applied to the system considered by performing a simple additional operation (e.g., applying a bit inversion substitution corresponding to either the encoded input bit vector or the encoded output bit vector). Whether or not the generation matrix includes bit inversion substitution does not affect the operation, characteristics, and effects according to embodiments of the present disclosure.

Encoded output bit sequence generated by the encoding is transmitted through the channel, and the receiver get alphabet here is the alphabet Unlike necessarily a binary field It doesn't have to be. codeword bit sequence Received vector from In addition to transmission to the channel, the transition involves various operations such as rate matching, rate dematching, interleaving and deinterleaving, modulation, and demodulation. It can be done. For simplicity of expression, we generally consider all these operations: at The vector channel between the ends of the This is taken into consideration.

The target of the decoder in the receiving device is the received vector Information bit vector from is to accurately estimate. The polar code decoding method and device are received vector can be used as is, or the corresponding log likelihood ratio (LLR) vector It can be converted and used. Bundles a series of processes of encoding and channel transmission at The vector channel of It can be expressed as: channel can be expressed as a vector channel obtained through channel combining.

The polar-signed SC decoder encodes the input bit vector is decoded sequentially one by one in ascending order of the index, and the estimated value is get In other words, the SC decoder is Decode the bits in order. The bits that have already been decoded and estimated are the received vectors Alternatively, it can be removed from the corresponding decoding target and used to facilitate decoding of the next bit. Because of these characteristics, the decoding operation may be referred to as sequential cancellation (SC).

By the operation of the SC decoding, the vector channel silver The channel is split into binary-input channels for each bit. The channel-separated binary-input channel may be referred to as a sub-channel, split channel, polarized channel, etc. Hereinafter, the expression side channel is used in this disclosure. In SC decoding second bit The decoding is the observation result of the given channel. and previously estimated It is performed based on Under the premise that the previously estimated bits are accurate (i.e. ), The side channel for is In [1], it is defined as [Equation 4].

[Equation 4]

Through the channel combining and separating process described above, side channel for This becomes polarization. That is, a set of side channels Some of them become excellent binary-input subchannels with a channel capacity close to 1, and others become bad binary-input subchannels with a channel capacity close to 0. index by side channel The channel capacity of is determined, and when designing a polar code, the channel capacity of the subchannel for each bit is deterministic. Therefore, the most important operating principle of the polar code system is In this case, information bits are allocated to the index with high channel capacity, and the values are fixed without allocating information bits to the remaining indices. A bit to which no information is assigned is called a frozen bit (or frozen bit), and its value is usually fixed to 0.

Encoded input vector for polar sign may be composed of a frozen bit (or fixed bit), a parity bit, and an information bit. The bits may be called by various other names. Typically, the parity bit is also called a dynamic frozen bit. Additionally, bits other than frozen bits are also called unfrozen bits. An information bit is a bit carrying information that can have a bit value of 0 or a bit value of 1, and the information bit can be assigned to a subchannel with a high channel capacity in polar codes. Fixed bits are bits with fixed values and can be assigned to subchannels with low channel capacity. The value of the fixed bit is usually determined to be 0, but this is not necessarily restricted (i.e., the value of the fixed bit may be determined to be 1). These fixed bits are bits used for channel polarization, and can be interpreted as fixing the bits to improve the channel capacity of other subchannels at the cost of giving up the amount of information. The parity bit in the polar code is the same as the fixed bit in that it does not carry information, but its value is not fixed to a specific value and is causally generated based on previous bits (bits with small indexes). The causally generated parity bit is used to improve error correction performance or error detection performance by using or checking the decoding result of the previously decoded and estimated information bits in the SC-based decoding method and device to be described later. According to one embodiment, the parity bit may be a cyclic redundancy check (CRC) bit. Additionally, according to one embodiment, the parity bit may further include a parity check (PC) bit defined in 3GPP.

length 3-bit information bits, given using this 8-pole symbol Let us consider an embodiment in which is encoded. Encoded input bit vector In each of the information bits It can be assigned and encoded. Corresponding encoded input bits Is Can be determined by processing with arbitrary permutation, in this embodiment , , It is assumed that mapping is done in index order as shown. The parity bit is 1 bit, and the parity bit is 1 bit. can be assigned to The parity bit can be generated by linearly combining bits with an index smaller than the index of the parity bit (e.g., 6). In this example It can be decided as follows. Here the symbol ' ' represents binary sum (modulo-2 sum). Bits other than information bits and parity bits is a fixed bit whose value can be fixed to 0. Likewise Deciding whether to use each bit as an information bit, a parity bit, or a fixed bit may be referred to as subchannel allocation, rate profiling, etc. As a result of the subchannel allocation described above, the encoding input bit vector is can be decided.

When representing the sign configuration of a polar sign, the set symbol of the bit index It can be used as sign is used as an index set of encoded input bits to which information bits are assigned, and in the above example, This happens. sign is used as an index set of encoded input bits to which parity bits are assigned, and in the above example, This happens. sign is used as an index set of encoded input bits to which fixed bits are assigned, and in the above example, This happens.

[Equation 4] for channel polarization is too complex to calculate as is, so it is practically impossible to calculate. second bit The number of additions required to decode one is approximately the set It is about the size of . for example, This is 64, 5 Ramen, set size of Is It becomes. This is not a level that can be handled no matter how high-performance hardware and excellent algorithms are used.

For realistic SC decoding, [1] showed that the transition probability defined in [Equation 4], or the equivalent LLR, can be calculated through a simple recursion operation. This repetitive operation is possible because of the generation matrix used to encode the polar sign, such as [Equation 2] and [Equation 3] is a small matrix This is because it is composed repetitively and regularly. Conventional SC-based decoding methods and devices have transition probability However, the value corresponding to the transition probability is not calculated using the operation described in [Equation 4]. Conventional SC-based decoding methods and devices calculate using a simple iterative operation based on the structure of the generation matrix. In this case, the amount of operations (number of additions and equivalent operations) required to perform SC decoding is It is suitable for implementation.

An embodiment of the SC decoding operation for a conventional polar code is described below. This conventional embodiment considers SC decoding using LLR. However, the content of the embodiments of the present disclosure is not limited to the LLR-based polar code decoding method and device, and can be applied to any type of SC-based decoding method and device.

First, to decode the polar code, the length The input LLR sequence of is given. SC-based decoding methods and devices encode input bit vectors It is characterized by decoding each bit sequentially, one bit at a time, according to the index order. In ascending order of the indices of the encoded input bit sequence, that is, Decoding is performed for each bit in order. Specifically, the SC-based decoding method and device operate as follows.

① Decode bit index is initially set to 0.

② If If is an information bit, the input LLR sequence and the estimated values of the previously decoded bits are Based on Calculate LLR, which is a probability-based metric for the 0s and 1s of Based on this value Estimation for Decide.

③ If If is a fixed bit or parity bit rather than an information bit, then by the corresponding method estimate of can be decided.

④ Estimation to decode the next bits is reflected in the decoder based on sequential cancellation (SC).

⑤ Decode bit index Increment by 1 and perform process ②. If the decode bit index go When this happens, decoding ends.

The encoding and decoding processes of polar codes can be understood and implemented based on belief-propagation on a bipartite graph corresponding to the generation matrix in FIG. 3.

Figure 3 shows an example of a graph corresponding to a polar sign in a communication system or broadcasting system according to embodiments of the present disclosure.

Referring to Figure 3, in Figure 3 the size The generation matrix of this 8-member polar sign is An example of a binary graph 310 for is shown. The binary graph consists of variable nodes (331) represented by circles, check nodes (332) represented by squares, and edges (333) connecting these nodes. . A variable node represents an individual bit, and a check node represents a linear constraint on the bits of a connected neighboring variable node. Specifically, the check node indicates that the binary sum (modulo-2 sum) of all bits corresponding to neighboring variable nodes is 0.

In a bipartite graph of polar signs, the connection between the variable node and the check node is the production matrix It is determined by composition. Generation matrix as shown in [Equation 2] and [Equation 3] is the polarized kernel Since it is generated with a Kronecker power of , the entire bipartite graph 310 has a repetitive and regularly connected form with the graph of the polarization kernel 320. Specifically, the polarization kernel 320 is a linear transformation for input size 2 and output size 2. It describes the relationship between and its shape represents a Z shape. generation matrix Is Since it is made with a continuous Kronecker power of , the polarization kernel 320 of the entire binary graph 310 is composed of a regular arrangement of a Z-shaped graph, as shown in FIG. 3.

Length configured as above The graph of the polar sign is It consists of stages. sign When using as the stage index of the graph, the stage index for the leftmost variable nodes of the graph is , the step index for the rightmost variable nodes in the graph. It is displayed as Among these, the leftmost step ( )of The variable nodes 340 are encoded input bit vectors. corresponds to the rightmost step ( )'s variable nodes 350 are encoded output bit vectors. corresponds to

The generation matrix shown in [Equation 1] The encoding process of polar codes by multiplication can be understood on the graph of FIG. 3. The leftmost step on the graph in Figure 3 ( )of Encoded input bit vector given to variable nodes 340 is assigned. And step index The bit values of the variable nodes are determined to satisfy the relational expression of the check node in ascending order (from left to right). In other words, the encoding process can be understood as an operation of updating bit values according to the relationship between variable nodes and check nodes sequentially from left to right in the graph shown in FIG. 3. Encoding by generation matrix multiplication of [Equation 1] can be implemented and realized by any method, and the method using the graph also produces the same encoding result.

SC-based decoding of polar codes can also be understood as a belief-propagation operation on the graph of FIG. 3. As the first step of decoding, the codeword bit vector (In reception, the reception vector ) for the rightmost step of the graph ( )of Corresponds to or inputs variable nodes (350). And through a series of actions, the leftmost step of the graph ( ), the encoded input bit vector corresponding to the variable node of The posterior LLR (a-posterior LLR, AP-LLR) of is calculated sequentially one by one. The AP-LLR of each bit is obtained by sequentially calculating and updating the LLR of the variable nodes for each step from the right side of the graph to the left side (i.e., in descending order of the step index). That is, the initial LLR is is given in the steps, from which The LLR values for some variable nodes in each step are calculated in order. through this process The AP-LLR value for the variable node of the step is calculated.

The detailed LLR calculation process on the polar sign graph is as follows. As described above, the entire bipartite graph is composed of the Z-shaped polarization kernel 320 as a basic element, and all calculations of log-likelihood ratio (LLR) values are performed on this basic element.

FIG. 4 illustrates an element process for performing log-likelihood ratio (LLR) calculation and sequential removal in a polar code decoding method and device for a communication system or broadcasting system according to an embodiment of the present disclosure.

Referring to Figure 4, the LLR calculation process on the basic elements is shown. Bit vectors corresponding to the left two variable nodes in this Z-shaped elementary element graph are , the bit vectors corresponding to the two variable nodes on the right are I decided to write: Polarized Kernel On by and The relationship can be expressed mathematically as follows.

[Equation 5]

beat The LLR values corresponding to Let's write: In decoding each basic element, the corresponding right two variable nodes are , is given, and by the decoding process , One LLR value is calculated. According to the sequential decoding method of the polar code described above, The LLR value of is calculated and estimated, based on this The LLR value of is calculated and estimated. first is the function in Figure 4 It can be calculated using (410).

[Equation 6]

[Equation 6] is a method of accurately calculating the LLR value according to the relationship between bits, and for simpler calculation and implementation, a function is used as in the equation below: It can be obtained by approximating .

[Equation 7]

In the above equation, the function is a function that returns the sign of the input value. If the input value is greater than 0, 1 is output, and if the input value is less than 0, -1 is output.

bits as decoding progresses. estimate of Once obtained or available, the function in Figure 4 via 420 Calculate .

[Equation 8]

As previously described, in the SC-based decoding method and device, bits are decoded and estimated sequentially according to the bit index. thus Decoding and estimation must be It is performed after confirming or estimating the value of .

The LLR calculation process described using FIG. 4 refers to the basic component of the polarization kernel, and this basic component may be connected to other variable nodes and check nodes as part of the overall graph. The default component is the leftmost step ( ), the calculated above class is the left side (i.e. the step index is passed on to other Z-shaped primitives connected to 1 (smaller side). And the steps are equally performed in the delivered basic components. In this way, decoding is done in steps In descending order of the step index, i.e. proceeds in decreasing order). The leftmost side of the graph (step When the LLR value of the encoding input bit of the variable node of ) is obtained, the bit value is estimated based on the LLR value. If the bit is a fixed bit, a pre-determined value (e.g., 0) can be substituted regardless of the calculated LLR value. If the corresponding bit is a parity bit, the value can be determined based on the previously estimated or determined bit. If the corresponding bit is an information bit, the hard-decision result of the updated LLR value may be substituted into the corresponding bit. Encoded input bit The final calculated AP-LLR for If you write, the decoder makes an estimate based on this value. Decide. For example, if LLR is defined so that positive numbers correspond to bit 0 and negative numbers correspond to bit 1, it is estimated as in the equation below: is decided.

[Equation 9]

In the case of floating-point operations, if the LLR value is 0, the bit value is determined according to a predetermined rule. The predetermined rule method may also be a random decision.

The bits estimated as a result of decoding are transferred from the left to the right of the graph to variable nodes in the range possible in the current decoding step, and are used to calculate the LLR of other variable nodes as in the process of [Equation 8]. This process can be understood as a sequential removal operation that reflects and removes the results decoded to date from the reception vector or LLR that is the target of decoding to facilitate decoding of the next bit. This sequential removal process is also a function of Figure 4. It can be understood from the basic components of the Z shape, such as (420) and SC (430). Decryption is in progress estimate of When obtained, this is the function we looked at earlier (420) and is reflected in the graph through [Equation 8]. Makes calculations possible. Decryption is in progress , estimate of , If obtained, this is reflected in the graph as SC 430. , The estimated person of , It makes calculation of possible. Specifically, according to the relationship of the polarization kernel , is decided. When the bit corresponding to the right variable node of the basic component is estimated, this value is again calculated on the right side of the graph (i.e., the step index is passed to other Z-shaped primitives connected to 1 (larger end). According to the above decoding procedure and process, the encoded input bit vector The bits are decoded one by one in ascending order of the index. The process of sequentially estimating and determining bits in the SC-based decoding process can be understood based on a binary decision tree.

FIG. 5 illustrates an example of a binary decision tree for sequentially decoding and estimating bits in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure.

Referring to Figure 5, in Figure 5 the size An example of a binary decision tree 500 for the process of sequentially estimating and determining the encoding input bits of this 8-pole code is shown. This embodiment uses the subchannel allocation discussed earlier, specifically encoding the input bit vector at are fixed bits, are the information bits, and Is This is a parity bit generated by .

SC-based decoding method and device Decode the bits in ascending order of index. first is a fixed bit, estimated is immediately determined to be 0. Therefore, as shown in Figure 5 No branching occurs for the value of . equally class It is also a fixed bit, and assumes class are all determined to be 0. beat is an information bit, estimated based on the LLR calculated through the above series of processes. is determined as either 0 or 1. Accordingly, a branch occurs in the binary decision tree. Fixed bit in the same way and information bits is decoded and an estimate for it is obtained. parity bit can be processed and decoded in two ways. first, Accordingly, based on the previously estimated values make an estimate for can be decided. Second, through the above series of fixations Calculate the LLR for and make a hard decision on it can be obtained. this estimate Based on the previously decoded result, It can be compared to see if it is equal. If the two results are different, the decoder may determine that the decoding result so far is wrong because the decoding result violates the given constraints. Based on this comparison information, the decoder can perform subsequent operations suitable for the decoding algorithm in use, such as pruning invalid paths or terminating decoding early. The binary decision tree in Figure 5 contains the previously decoded and valid depending on the value Only the decision path is shown. chamberlain ego If so, Since this cannot be done, the judgment path for this is excluded.

The above process is about the SC decoding method and device, and this process can be used with some modification in various variations such as SCL decoding, SCS decoding, and SCF decoding.

Now, specific embodiments of the present disclosure that identify and overcome the limitations of conventional SC-based decoding method devices are described.

The transition probability calculated by conventional SC-based decoding methods and devices is ultimately based on [Equation 4]. That is, if the operations used in the recursive operation of the SC-based decoding method and device are not approximated, the final calculated value is the same as the value calculated by [Equation 4]. For LLR-based decoding methods and devices, encoded input bits as in [Equation 9] AP-LLR calculated for If written, this value is the same as the following equation based on [Equation 4].

[Equation 10]

Here, we used the definition that a positive LLR value indicates bit 0 and a negative number indicates bit 1. Even if this definition is opposite, it can be understood and implemented by changing the signs of LLRs generated and calculated in the decoding method and device.

Looking at the right side of [Equation 4], which is the basis of the conventional SC-based decoding method and device, The transition probability of the subchannel for all bits after All values that can have (i.e. ) is the result of marginization. However, due to the composition of the extreme sign, There may be constrained bits such as fixed bits or parity bits. The existence of these restricted bits shows that the transition probability of the subchannel in [Equation 4] is defined and is not accurate when considering the actual configuration of the polar code.

In other words, in conventional SC-based decoding and devices The transition probability and LLR calculated for Decoded after viewing does not take into account the composition of The calculation method of the LLR is All bits that are decoded later are treated as pure noise. Despite this relaxation, as demonstrated in [1], the polar code can achieve channel capacity as its length increases. However, this simplification shows that the conventional practical SC-based decoding and apparatus for finite-length polar codes are not optimal because they do not consider constraints on bit values. Hereinafter, the above content is visually described through examples.

FIG. 6 illustrates an example of a binary decision tree considered in a polar code decoding method and apparatus of a communication system or broadcasting system according to embodiments of the present disclosure.

Referring to Figure 6, the binary decision tree 600 is an information bit in the embodiment seen in Figure 5. When performing decoding for , an example of a binary decision tree considered in a conventional SC-based decoding method and device is shown. Conventional methods and devices are In decoding Assume that all bits of are not constrained. The sub-tree 610 of the binary decision tree 600 in Figure 6 is It contains paths for all 16 combinations that the 4 bits of can have. However, considering the composition of the polar sign, is a fixed bit and Since is a parity bit, each About value The actual number of possible combinations that can be had is less than 16 combinations. Binary decision tree 601 is an information bit When performing decoding for , an example of a binary decision tree derived based on possible combinations is shown. As shown in sub-tree 620 of binary decision tree 601, there are only four possible combinations. Therefore, the binary decision tree 600 considered in the conventional method and device is different from the binary decision tree 601 according to the actual code configuration, which shows the sub-optimality of the conventional decoding method and device. .

The decoder according to embodiments of the present disclosure to solve the above problem uses the conventional [Equation 4] The following equation By accurately modifying it as follows, decoding can be performed.

[Equation 11]

In [Equation 11] above, was previously decrypted and is being considered under the condition Given (i.e., combined with ), It is the set of all combinations that can be had. For example, in the example covered in Figure 5 If you decode is the same as shown in the sub-tree 620 of FIG. 6 and is specifically the same as the following equation.

[Equation 12]

As shown in [Equation 11] above, the transition probability that accurately considers the sign configuration is used in embodiments of the present disclosure. The LLR-based SC decoding and its modified method and apparatus according to embodiments of the present disclosure calculate the LLR defined as in the following equation.

[Equation 13]

A decoder according to another embodiment may use an LLR defined in [Equation 13] where the numerator and denominator are exchanged. The method using Equation 13 can be understood, implemented, and realized as inverting the sign of LLR.

An embodiment of the SC-based decoding method and device of the present disclosure, unlike the conventional method and device, is based on [Equation 11] that sets a margination set based on the actual code configuration. In this respect, the SC-based decoding method and device of the present disclosure may be referred to as bitwise maximum likelihood-successive cancellation (bitwise ML-SC) decoding. In other words, in the above-described embodiment of the present disclosure, the decoder can sequentially decode bits one by one based on sequential removal. The decoder can perform ML decoding that accurately reflects the configuration of the code for each bit.

Figure 7 shows the performance of a decoding technique according to embodiments of the present disclosure. In Figure 7, the performance of three decoding techniques for the example discussed in Figure 5 (pole code of length 8 consisting of 3 information bits, 1 parity bit, and 4 fixed bits) is shown.

Referring to FIG. 7, a graph 700 represents the performance of a decoding technique according to embodiments of the present disclosure. The vertical axis 701 represents block error rate (BLER), and the horizontal axis 702 represents signal-to-noise ratio (SNR). The first line 710 represents performance according to the Blockwise ML technique. The second line 720 represents performance according to the SC decoding technique. The third line 730 represents performance according to the Bitwise ML-SC technique.

“SC” decoding is a conventional method based on [Equation 4]. “Bitwise ML-SC” decoding is a method realized/implemented/operated based on [Equation 11] according to an embodiment of the present invention. In addition, “Blockwise ML” decoding, which is the optimal decoding method for linear codes such as polar codes, was also considered. The Blockwise ML method is an exhaustive search method that compares a given received vector with all valid codeword vectors and selects the most likely one. The codeword vector generated in this experiment is modulated with binary phase shift keying (BPSK) and transmitted through an additive white Gaussian noise (AWGN) channel. In this environment, the graph 700 provides a block error rate (BLER) result by comparing the encoding input bit vector and the decoding result when each of the above three methods is used. is a parameter representing SNR and is used as a quantitative metric of channel quality.

As shown in FIG. 7, conventional SC decoding has lower performance than optimal Blockwise ML decoding due to sub-optimality in [Equation 4]. Low performance is the same A higher BLER means more block errors occur. Additionally, lower performance requires higher signal strength to achieve the same block error rate, i.e. This means it is necessary. On the other hand, Bitwise ML-SC decoding according to an embodiment of the present disclosure based on [Equation 11] provides the same performance as optimal Blockwise ML decoding on a block basis in this example. Through this, the non-optimality of the conventional method and the optimality of the present invention can be confirmed experimentally.

We now describe how to effectively implement the SC decoding method and device based on [Equation 11]. As explained earlier, calculating [Equation 4] as is requires a very large number of operations. In [1], it was shown that [Equation 4] can be implemented with low complexity through recursive operations on polarized kernels, based on the structure of the generation matrix. Likewise, calculating [Equation 11] as is also requires a very large number of calculations, and the present invention includes a method of effectively realizing/implementing and operating the above process.

In one embodiment of the present disclosure, sequential cancellation check (SCC) decoding is used to realistically calculate the transition probability and LLR according to [Equation 11]. In the conventional SC-based decoding method and device, the decoder uses information bits (in other words, When decoding ), calculate metrics such as transition probability or LLR for this estimate of get In one embodiment of the present disclosure, first the information bit When decoding, this value can be substituted or assumed as 0 or 1, respectively. next, At least one of the bits located behind (in other words, ), the decoder can check or evaluate the validity of the substitution or assumption. And, the decoder estimates based on the above inspection or evaluation. can be decided.

That is, the SCC decoding method and device of the present invention contain at least one information bit ( ), the estimated value of The bit located behind ( ) is processed through a series of processes and then decided. especially, may be a fixed bit or a parity bit. The fixed bit or parity bit with a larger index than the bit to be decoded is called a future constraint, future constrained bit, post-constraint, or post-constrained bit. ), etc. remind The future constraint bits considered for the estimation of may be one or two or more.

One embodiment of SCC decoding may consist of the following series of processes.

- The decoder uses information bits (in other words, ), the target bit index (target bit index) Select . Is An integer greater than or equal to is selected. For example, Is While can be chosen as the smallest integer. in other words, may be the bit immediately preceding the next information bit. in this case, are all fixed bits or causally generated parity bits, so by definition, the preceding bits If given, its value can be determined immediately.

- you want to decrypt Assume the values of are 0 or 1, respectively. The decoder previously decrypted and as assumed above based on Decide. As a result, the decoder has a total Check .

n symbol Assume based on (or ) means a decided hypothesis.

n symbol Assume based on (or ) means a decided hypothesis.

- The decoder guesses each and based on Perform decoding. In an embodiment of this decoding process, the decoder You can calculate a probability-based metric for and select at least one guess based on it. In another embodiment, the decoder At least one guess can be selected by calculating a probability-based metric for . In another embodiment, the decoder At least one guess can be selected by calculating a probability-based metric for . In another embodiment, the decoder You can check and check the validity of and select at least one guess based on this. In another embodiment, the decoder You can check and check the validity of and select at least one guess based on this. In another embodiment, the decoder You can check and check the validity of and select at least one guess based on this.

As described above, the principles of the present disclosure can be realized and implemented in various ways. An important point in embodiments of the present disclosure is that at least one information bit Make estimation decisions and judgments about This means that it can be performed after processing at least one other bit located behind it. According to one embodiment, a decoder using polar codes can estimate information bits corresponding to a specific bit index after margining using a freeze bit or parity bit located behind the specific bit index. .

The SCC decoding method can be expanded and applied to SCL decoding, SCS decoding, etc., which are modifications of the conventional SC decoding. In other words, the concept of the SCC decoding method can be applied to a method that considers two or more candidate groups rather than leaving only one candidate group in the sequential decoding and bit determination process. For example, during the decision process of each bit, the maximum The concept of SCC decoding can be applied to SCL decoding, which maintains a group of provisional candidates called a list. The technique in which the SCC decoding is applied to SCL decoding may be referred to as SCCL (SCC-list) decoding. Information bits in the SCL decoding When decoding, the decoding result of the previous bit is list of can be given. with a list of With a combination of assumptions about the values 0 and 1, total of A group of candidates can be determined. Specifically class about, The second group of candidates is class It can be decided based on . This candidate group is class decided based on may include. The decoder may perform the SCC decoding based on the candidate group. That is, in decoding each list, The bit estimation of is person It can be determined during or after processing. At this time, a path metric (PM) used in conventional SCL decoding may be used. As a result of decoding, the total From the candidate group of You can select a list and proceed with the subsequent process. In the same way, the SCC decoding method can be extended and applied to SCS decoding, and in this case, it can be implemented and realized as SCCS (SCC-stack) decoding.

The above SCC decoding method and procedure partially supports the concept of [Equation 11].

FIG. 8 shows an example of a binary decision tree used in a polar code decoding method and apparatus of a communication system or broadcasting system according to embodiments of the present disclosure.

Referring to Figure 8, binary decision tree 800 illustrates an example of a binary decision tree considered by the SCC decoding method and apparatus in the running example above. The decoder uses information bits Based on the above SCC decoding method, estimate the next information bit Future pharmaceutical bit in front It can be performed in . in this case, The estimate for is fixed bit This is done after the exact value of is reflected in decoding. In this way, the SCC decoding method and device have the characteristic of reflecting some future constraint bits close to them when decoding and estimating information bits. This is a feature of an embodiment of the present invention that distinguishes it from conventional SC-based decoding methods and devices.

In another embodiment of the present disclosure, belief-propagation successive-cancellation (BP-SC) decoding can be used to realistically calculate the transition probability and LLR according to [Equation 11]. there is. The BP-SC decoding method and device can use the polar code graph introduced in FIG. 3 by modifying it based on future constraints. The decoding method and device using polar codes according to embodiments of the present disclosure can utilize future constraints that were not considered in the conventional SC-based decoding method and device.

The process of reflecting future constraints on the bipartite graph used in decoding is as follows. The encoded input bit vector Constraints by the fixed bits and parity bits of the parity-check matrix It can be expressed in the following mathematical equation.

[Equation 14]

The parity check matrix of the first column It is expressed as At this time, parity check matrix Each column of is a encoded input bit vector Depending on the bit configuration, it is structured as follows.

- If is an information bit, then the parity check matrix of All elements in the first column consist of 0. That is, the heat vector is the 0 vector.

- If is a freeze bit (or fixed bit), the parity check matrix of second column Only the element in the first row has the value 1, and all other elements have the value 0. heat vector Is It is a vector where only the first row element is 1 and the rest are 0. For example, the road I keep looking at Example of polar sign ( , , ) fixed bit The column vector of the parity check matrix by It is given as follows.

- If is a causally generated parity bit, then the parity check matrix of The third column is configured to correspond to the parity check equation. For example, the road I keep looking at Example of polar sign ( , , ) in the parity bit go If generated by the relational expression, the column vector is It is given as follows.

As above, the configuration of the polar sign (encoded input bit vector ), the decoder uses the parity check matrix The configuration can be confirmed or determined.

For example, the continuation length above Example of polar sign ( , , ), the encoded input bit vector in Parity check matrix for can be given by the above rule as follows.

[Equation 15]

Generation matrix given in [Equation 2] and [Equation 3] above Is It is a self-inverse matrix, or an involutory matrix. in other words, has a relationship. Through this, [Equation 1] can be expressed as the following equation.

[Equation 16]

Based on [Equation 14] and [Equation 16] above, the equation can be defined as follows.

[Equation 17]

[Equation 17] above is On by From the constraints of On by It shows that constraints can be derived. In other words, the encoded input bit vector Parity check matrix for from the codeword vector Parity check matrix for can confirm. For simplicity of expression, codeword vector The parity check matrix for I decided to write: For example, the continuation length above Example of polar sign ( , , ) codeword vector in Parity check matrix for can be given by [Equation 17] as follows.

[Equation 18]

Above, shown in [Equation 14] and [Equation 17] and Constraints on can be expressed on a polar sign graph. Hereinafter, through Figure 9 and The constraints are described.

FIG. 9 illustrates an example of a graph modified in consideration of future constraints in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure.

Referring to Figure 9, the continued length Example of polar sign ( , , ) appears in and Constraints on can be expressed on the polar sign graph 900. Constraint 910 is in the above example [Equation 15] On by The five constraints (4 fixation bits, 1 parity bit) are indicated by connection lines related to the test node. For example, parity bit Constraints on 3 variable nodes in a graph It is displayed as a test node connecting . Constraint 920 is in the above example [Equation 18] On by The five constraints above are indicated by connection lines related to the test nodes. For example, constrained by [Equation 17] silver Same as, 4 variable nodes in the polar sign graph 900 It is displayed as test nodes connecting .

In the present disclosure, the decoder uses a codeword vector By checking the constraints on , decoding based on at least one future constraint can be performed. Conventional SC-based decoding methods and devices are Because the bits of are decoded sequentially, The future constraint bit located in could not be considered structurally. For example, In decoding, parity bit Constraints due to were not considered in conventional SC-based decoding methods and devices. On the other hand, the decoder according to embodiments of the present disclosure, as confirmed through [Equation 17] above, The codeword vector for future constraints located at It can be converted to the corresponding constraint in . The decoder according to embodiments of the present disclosure can improve decoding accuracy by performing decoding based on the result of the transformation. As shown in Figure 9, the codeword vector The restrictions on Can be used to decode each bit of.

FIG. 10 illustrates examples of graphs considering future constraints configured when decoding each bit in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure. The SC-based decoding method and device according to embodiments of the present disclosure can use a graph with constraints as shown in FIG. 9 interchangeably when processing each bit.

Referring to Figure 10, the continued example ( , , ) from the information bit A graph 1000 constructed when decoding is displayed. Information bits in SC-based decryption method and device When decoding, the previously estimated This is reflected as variable node 1001, variable node 1002, and variable node 1003. Therefore, already Fixed bit reflected in, corresponding to parity bit Constraints in can be excluded from the graph. In other words, In the decoding of , the future constraint bits are and and the corresponding Only the constraints in can be considered, such as constraints (1011) and constraints (1012).

In addition, the SC-based decoding method and device of the present disclosure necessarily apply future constraints to the rightmost step ( ) located in It can be used in the subcode of the polar code without considering it above. Hereinafter, through FIG. 11, an example of a decoding process in which future constraints are reflected in an intermediate stage rather than the rightmost stage is described.

FIG. 11 illustrates examples of graphs considering future constraints configured when decoding each bit in a polar code decoding method and apparatus for a communication system or broadcasting system according to embodiments of the present disclosure. The example continues above ( , , ) from the information bit In decoding, the future constraints are Parity bit generated by am. By sequential SC decoding estimated value of is available, so the above constraint is It can be expressed again as However, the above constraint formula at and All belong to the second sub-code in the repetitive and stepwise composition of the polar code. Here, the second sub-sign is length generated by It is the sign of .

Referring to Figure 11, the continued example ( , , ) from the information bit A graph 1100 constructed when decoding is displayed. The decoder enters the variable area 1110 (step ) generates a codeword corresponding to the intermediate variable node in Therefore, the formula by future constraints silver stage in Rather than, step as shown in the inspection node 1120 of Figure 11 in It can be converted and applied. The decoder for the SC-based decoding method and device according to embodiments of the present disclosure can determine the location of a newly added check node according to future constraints by reflecting the structure of the sub-code.

Like the above embodiments, the decoding method and apparatus of the present disclosure perform operations based on a graph of polar codes modified in consideration of future constraints. The transformation of the polar sign is at least one of the general transformation methods for a bipartite graph of a linear sign, such as adding a check node and connection line considering future constraints, deleting a node and a connection line, merging two or more nodes, and deleting a duplicate connection line. Includes one.

Decoding of polar codes on the modified graph can be performed based on belief-propagation. In conventional SC decoding, the calculation of LLR is unidirectional on the right side of the graph ( ) to the left ( ) was carried out only. However, as explained earlier, as new check nodes are added, bidirectional LLR calculation becomes possible in the graph. This process is the same as the message passing decoding of the low-density parity-check (LDPC) code or a modified method thereof. Repeated exchange of messages is also possible on this graph. In this respect, the method of sequentially performing decoding bit by bit through trust-propagation on the modified graph may be referred to as BP-SC decoding.

The two embodiments of SCC decoding and BP-SC decoding can be used in combination. The SCC decoding is characterized in that decoding is reserved to reflect some future constraints in the sequential decoding and bit decision process, and the BP-SC decoding is characterized in that the decoding graph used is modified based on future constraints. . The two methods are not mutually exclusive and the two methods can be used together. For example, the SCC decoding and BP-SC decoding can be combined to implement the BP-SCC decoding method.

Additionally, the embodiments of the present disclosure are not exclusive to SCL decoding, SCS decoding, and SCF decoding, which are extension methods of conventional SC decoding, and can be used to extend and apply the embodiments of the present disclosure to these methods. For example, the SCC decoding, BP-SC decoding, and BP-SCC decoding methods can be combined with methods such as list, stack, and flip that consider two or more candidates. Accordingly, other embodiments of the present disclosure include SCC-list, SCC-stack, SCC-flip, BP-SC-list, BP-SC-stack, BP-SC-flip, BP-SCC-list, BP-SCC-stack , BP-SCC-flip, etc. can be realized.

The decoding methods of the present disclosure can be implemented by solving a constraint satisfaction problem (CSP) by utilizing future constraints in a search process in a binary decision tree. As previously explained, SC decoding of polar codes can be understood as DFS (depth-first-search) that proceeds to child nodes for the estimated value for each bit in a binary decision tree. Conventional SC-based decoding selects one with a high probability among child nodes and performs a search based on metrics such as LLR. On the other hand, the newly added check node in the modified graph considering the future constraints can be used to determine the validity of the intermediate results obtained in the trust-propagation process. Therefore, the SC-based decoding method and device utilizing the future constraints of the present disclosure checks whether the future constraints are satisfied using the SCC, BP-SC, BP-SCC, etc. method, and then searches for a binary decision tree based on this. You can proceed.

The search of the binary decision tree for the SC-based decoding of the present disclosure may be performed in the manner of an embodiment as follows.

- If only one of the nodes satisfies the given future constraints (or does not violate the future constraints), the decoder proceeds with the search to that node.

- If two or more nodes satisfy existing constraints (or do not violate future constraints), the decoder can select one node according to a predetermined rule and proceed with the search. The predetermined rule may be a method of selecting one node based on a metric calculated or derived during the decoding process. The predetermined rule may be a method of randomly selecting one of the nodes. The predetermined rule may be a method of selecting a fixed node among nodes. The locations and information of nodes that are not selected while satisfying the future constraints (or without violating the future constraints) according to the above rules can be stored and managed in memory, etc., and these nodes can be used to prevent any nodes from subjecting the future constraints in subsequent searches. If it is not met (or if all nodes violate future constraints), it may be visited or exploited.

- If all child nodes do not meet a given future constraint (or violate a future constraint), according to one embodiment, the decoder may terminate the search and terminate decoding early. In another embodiment of the present disclosure, the decoder may perform backtracking to nodes that met future constraints (or did not violate future constraints) but did not proceed or visit during the previous search process. This process can be understood as conflict-directed backjumping (CBJ). In this process, one of the nodes stored in the memory, etc. can be selected according to a set rule. The set rule may be a method of randomly selecting one of the nodes stored in memory, etc. Additionally, the above-described rule may be a method of implementing memory in a stack structure and selecting the closest node among the nodes currently being searched.

The above descriptions are implementation examples of using future constraints according to embodiments of the present disclosure when SC-based decoding. It should be noted that the embodiments of the present disclosure can be implemented in any way that utilizes future constraints in SC-based decoding of polar codes and are not limited in any particular way.

Figure 12 shows the operational flow of a device for decoding based on future constraints according to embodiments of the present disclosure. The device may include a decoder. In Figure 12, a receiving device that receives a signal is described as a decoder, but this term is not to be interpreted as excluding the transmission or encoding operation of the receiving device. Of course, the receiving device can operate as a transmitting device in relationship with other devices.

Referring to FIG. 12, in operation 1201, the device may receive signals encoded with polar codes. The device may be a receiving device. The receiving device can receive a signal encoded with a polar code from the transmitting device. The signal may include a bitstream. The signal may include bit strings. A signal may include multiple bits. According to one embodiment, the signal may be a signal encoded based on polar signs. The signal may contain information bits. In order to be encoded with a size of N according to a specified condition (N=2 ⁿ ), freeze bits may be used in addition to information bits. The signal may include freeze bits. For error detection during the decoding process, parity bits can be used. The signal may include parity bits. The transmitting device can transmit a signal encoded with a polar code by sorting subchannels based on channel capacity (or reliability) and allocating information bits to channels with relatively large channel capacity (or reliability).

At operation 1203, the device may identify a polar code configuration including information bits and parity bits. According to one embodiment, the polar code configuration may include bit index information. Bit index information represents a bit index set of bits of a received signal. Bit index information may indicate the index of each information bit. For example, bit index information is an index set of encoded input bits to which information bits are assigned. can represent. Bit index information may indicate the index of each parity bit. For example, bit index information is an index set of encoded input bits to which parity bits are assigned. can represent. For example, bit index information is a set of indices of encoding input bits to which freeze bits are assigned. can represent.

According to one embodiment, the polar code configuration may include calculation method information. Parity bits can be used for error detection. Parity bits can be determined by specific operations on other bits. According to one embodiment, the parity bits of the polar sign may be generated causally. For example, the parity bit can be determined based on a linear combination of other bits. As an example, in the example shown in Figure 5, the parity bit Is and The operation of the linear combination of (i.e. ) is determined by. The operation method information includes index information (e.g., 3, 5) or operation method (e.g., 3, 5) for the other bits. ) can refer to at least one of:

According to one embodiment, unlike what was described above, the signal may not include a parity bit. Since the parity bit is used in concatenated polar codes, it may not be used in certain polar codes. Accordingly, embodiments of the present disclosure may include both an operation of decoding a signal including a parity bit and an operation of decoding a signal not including a parity bit. According to one embodiment, if the signal does not include a parity bit, that is, if the parity bit is not used for channel coding, the polar code configuration may not include the parity bit.

At operation 1205, the device may perform successful cancellation (SC) decoding (i.e., sequential cancellation decoding). SC decoding refers to the operation of sequentially decoding one bit at a time based on previously estimated bits. The device can remove bits for which estimation has been completed from the bit stream and sequentially decode bit by bit starting from the remaining bits. To decode a signal encoded using a polar code, SC decoding can be used.

According to embodiments, the operation of the device performing SC decoding may include operation 1206. In operation 1206, decoding of the information bits may be performed based on constraints determined by at least one of a parity bit or a freeze bit corresponding to an order after the information bit. Here, the parity bit or freeze bit corresponding to the order after the information bit may be referred to as a future constraint (or future constrained bit). Hereinafter, decoding of the information bits may be performed based on at least one value of at least one constraint bit.

Decoding of the information bit may be performed based on the value of the bit located after the index of the information bit. In performing SC decoding, future constraints on parity bits and/or freeze bits according to embodiments of the present disclosure may be used to estimate information bits. Here, the parity bit and/or freeze bit are located after the information bit to be estimated. As an example, when bits are indexed in ascending order, the index of the parity bit and/or freeze bit may be larger than the index of the information bit.

The freeze bit has a fixed value. For example, the values of the freeze bits of the signal may all be 0. As another example, the values of the freeze bits of the signal may all be 1. The parity bit has a value determined according to the specified method. For example, the value of the parity bit can be determined through operations on information bits. In this way, freeze bits and parity bits are limited based on rules according to polar code configuration.

In SC decoding, the posterior-probability of the information bits is used. The device according to the embodiments of the present disclosure does not consider all bit values, but rather determines in advance the value of the constraint bit (i.e., the freeze bit or parity bit corresponding to the subsequent decoding order of the information bit) and then decodes the information bit. can be performed. Since the post-probability calculation is used only for certain candidates, not all candidates (i.e., marginization in [Equation 11]), the decoding results of information bits can be more accurate.

Although SC decoding has been described as an example, it goes without saying that the decoding principle according to the future-constraints of the present disclosure can be applied to other decoding techniques other than SC decoding.

Hereinafter, the decoding technique using the future constraint described above with reference to FIGS. 1 to 12 will be described in more detail through equations. For this purpose, the following additional mathematical signs and symbols are defined and used for vectors and matrices.

- As explained previously, unless otherwise specified, the index of the first element of a vector, matrix, or sequence is assumed to start from 0.

- Two non-negative integers About, Is from Represents a set of consecutive integers up to. When this symbol is used as a subscript to indicate a set of vector and matrix elements, the parentheses are omitted for more concise expression. It can be used as

- Vector and two non-negative integers About, is a vector of From the first element Represents a subvector consisting of up to the th element. in other words, , and in other notations, It can also be written as if go If it is less than is a NULL vector with size 0.

- procession and two non-negative integers About, is a matrix of second row, Indicates the element of the th column.

- procession and a set of two non-negative integers About, is a matrix The index from The rows and indices belonging to Represents a partial matrix composed of columns belonging to . The above definition is intended to represent slicing of a matrix.

- For simple matrix slicing notation, the symbol Use . This symbol is used as a shorthand symbol to represent the set of all rows or all columns of a given matrix. In other words, the size phosphorus matrix About and represents.

The content of the present invention will be described in more detail based on the above additional mathematical symbol definitions.

FIG. 13 illustrates an example 1300 of a concatenated polar coding system using outer coding according to embodiments of the present disclosure.

Referring to Figure 13, the given information vector (information vector) is first encoded by outer encoding, and is called an outer codeword vector. creates . The bits of the external codeword vector have a length of Encoder input vector of is mapped to each location. Here, the bits of the external codeword vector are mapped to the position corresponding to the subchannel with high reliability (or high channel capacity) among the subchannels formed by channel polarization of the polar code, and the remaining bits The values are fixed as frozen bits. Afterwards, the encoding input vector is the generation matrix is multiplied, and the result of the multiplication is an encoder output vector or codeword vector. is created.

FIG. 14 shows an example 1400 of an equivalent model of a concatenated polar code system using external coding according to embodiments of the present disclosure. The concatenated polar code system including external encoding, described in FIG. 13, can be represented by the equivalent model of FIG. 14. Given information vector First, the length is determined through a process called rate profiling. extended vector of is mapped to Here, bit mapping can be performed considering the characteristics and performance of the subchannel by channel differentiation. Extension vector, the output of rate profiling By performing precoding on the encoding input vector can be created. Vector expansion by precoding A precoding matrix By multiplying with , the encoded input vector is created. For example, the encoding input vector is generated as shown in the following equation.

[Equation 19]

In the precoding process, one or more parity bits may be causally generated by external encoding. The subsequent process is the same as the process in FIG. 13 above.

That is, the equivalent model 1400 of FIG. 14 is a combination of outer encoding and sub-channel allocation and a combination of rate profiling and precoding in the process of FIG. 13. It is displayed by changing it to . Since each of the two system models produces the same output for the same input, the two system models are equivalent. However, the equivalent model 1400 of FIG. 14 (i.e., rate profiling and precoding model) is advantageous for explaining embodiments of the present disclosure in that external encoding can be expressed as a matrix product. Below, embodiments of the present disclosure are described in detail using the equivalent model 1400 of FIG. 14.

As explained previously, the parity bit is encoded in the input vector by external encoding and subchannel allocation. It can be causally created and placed on the computer. Considering this process, each column of the precoding matrix is structured as follows according to the type of the corresponding subchannel.

- Index corresponding to the information bit about, the first column is , person About It consists of

- Index corresponding to the fixed bit about, All elements in the first column are 0, i.e. It consists of

- Index corresponding to parity bit about, silver parity bit go It is constructed to indicate something that is causally created by . also is constructed so that all elements are 0.

For example, the continuation length above Example of polar sign ( , , ) precoding matrix in is given by the above rule as follows.

[Equation 20]

The entire generator matrix composed of a combination of precoding and polar coding is It can be. For example, the overall production matrix above is given by:

[Equation 21]

precoding matrix and the parity check matrix described above. By definition, when multiplied together, it becomes a matrix in which all elements are 0. For example, based on [Equation 19], the relationship of the following equation is established.

[Equation 22]

The parity check matrix Since the column for the information bits consists of all 0s as described, Equation 22 ( )Is It can be displayed in a simpler form as follows. for teeth A simple form such as can be defined and used.

Now, the SCC decoding process described above will be explained in more detail based on the formula. Information bits by sequential decoding Let's consider the steps to decrypt. As previously explained, the SCC decoder device is Rather than directly estimating , future constraint bits between the next information bits can be reflected. Bit to be estimated for effective distinction of bits is referred to as a target bit, and the bit to be actually processed may be referred to as a processing bit. target beat The index of the processing bit for It is written, and this value is determined according to the equation below.

[Equation 23]

if If so, the information bits are arranged in succession, and the device can perform general SC decoding because there are no future constraints to consider. if Ramen, Goal Beat The hypothesis can be established that the values of are 0 and 1, respectively. target beat The value of The assumption that It is decided to write with the symbol . home For verification of length Temporal vector of can be created. Previously estimated vector Temporary vector based on Can be generated sequentially by the following equation.

[Equation 24]

The above temporary vector The configuration can be created based on the property that parity bits are causally generated.

A device for SCC decoding according to embodiments uses two temporary vectors obtained through the same process as [Equation 24] above. class Based on , each SC decoding can be performed. The decoding process may be accomplished by calculating a metric for probability or determining validity when reflecting each temporary vector. Through this series of processes, the SCC decoding process can be interpreted as performing decoding based on improved transition probability as shown in the following equation.

[Equation 25]

channel Obtained through channel combining, at refers to the vector channel of Is represents a side channel for .

Now, the future constraint conversion process for BP-SC and BP-SCC decoding described above will be explained in detail based on mathematical equations. As previously explained, the encoded input bit vector is the parity check matrix About is limited by the relationship. Also the encoded output vector is created with Using these two relationships, an equivalent parity-check matrix is created for the encoded output vector. is defined by the following equation.

[Equation 26]

Based on this result, the parity check equation for the encoded output vector can be obtained as shown in the following equation.

[Equation 27]

The above parity check equation is the generation matrix is an involutory matrix, that is, It is obtained due to the phosphorus properties.

continuous length Example of polar sign ( , , ), the parity check matrix for the encoded output vector in is given as follows:

[Equation 28]

When decoding each information bit, the parity check equation [Equation 27] can be modified and used. First, random bits an index set of future constraint bits (which do not have to be information bits) It is written as , and is defined as follows.

[Equation 29]

information bit The modified parity check matrix for decoding is Let's write ., which is given as the following equation.

[Equation 30]

Based on the above, the information bit The parity check equation constructed when decoding is structured as follows.

[Equation 31]

SCC decoding is used to determine the target bit bits processed for estimation of If dealing with , the above [Equation 31] can be rewritten as the following equation.

[Equation 32]

The parity check equation given as [Equation 31] and [Equation 32] above is the target bit. It can be used directly for decoding. The first term of each equation is composed of previously estimated bits and hypothesized bits and can be obtained directly. The second term consists of the transformed encoded output vector of the future constraints. Since the encoded output vector directly corresponds to the decoder input (LLR vector) in the decoder, the device can immediately utilize the relationship. Therefore, since all terms can be used, the device can improve decoding performance by utilizing [Equation 31] and [Equation 32].

Now, we will explain the subgraph-based future constraint transformation in more detail through formulas. For this purpose, several symbols are additionally defined. First encode the input bits Sizes included The starting index of the subgraph is It can be expressed as Based on this, the index set of all bits belonging to the corresponding subgraph can be defined by the following equation.

[Equation 33]

By the above definition, the relationship of the overall polar sign Sizes included in The subgraph relationship of can be confirmed.

Figure 15 shows an example of an intermediate codeword vector in embodiments of the present disclosure. first is the encoded input bit Sizes included Represents an intermediate codeword vector generated by a subgraph of .

Referring to FIG. 15, the graph 1500 in which the above example is reflected may include a subgraph 1510 and a subgraph 1520. The subgraph 1510 is represents a subgraph of size 2 containing in vector(1511) can be obtained. Additionally, in Figure 15, the subgraph 1520 is represents a subgraph of size 4 containing in vector(1521) can be obtained. The following relational expression is obtained by recursive generation of polar signs.

[Equation 34]

is the encoded input bit Sizes included Represents the intermediate level vector generated by the subgraph of . vector and a set of non-negative integers About, Is The indices are set from the elements of Represents a subvector composed of elements belonging to . in other words, represents. in other words, , encoded input bits Sizes included The encoding of the index set of all bits belonging to the subgraph of the input bit vector, i.e., the subvector of the input vector. As explained earlier, Is of As a result of the Kronecker power, size It is the same as the polar sign generation matrix of .

Additionally, two index sets are defined as shown in the equations below.

[Equation 35]

[Equation 36]

Based on the definition of the set above, encoded input bits Sizes included An intermediate codeword vector generated by a subgraph of The parity check matrix for is defined as the following equation.

[Equation 37]

Based on this, the encoded input bit Sizes included The converted parity check equation for the subgraph can be obtained as in the following equation.

[Equation 38]

SCC decoding is used to determine the target bit bits processed for estimation of If dealing with , the above [Equation 38] can be rewritten as the following equation.

[Equation 39]

As with other parity check equations examined previously, all terms in [Equation 38] and [Equation 39] are target bits. It can be confirmed when decoding. The decoder of the device can improve decoding performance based on the confirmation result.

According to embodiments of the present disclosure, a method performed by a device in a communication system or broadcasting system may include receiving a signal including bits encoded based on polar codes. The method may include identifying a polar code configuration for at least one information bit and at least one frozen bit of the signal. The method may include performing decoding of the signal based on the configuration of the polar code. Decoding of an information bit among the at least one information bit may be performed based on at least one value of at least one constraint bit. The at least one constraint bit may include at least one of a parity bit or a freeze bit corresponding to a later order than the information bit.

According to one embodiment, each bit of the at least one constraint bit may be included in at least one parity bit or the at least one freeze bit of the signal.

According to one embodiment, the operation of performing the decoding includes determining the value of a constraint bit corresponding to a subsequent information bit based on a previous decoding result of the information bit and candidate values of the information bit. can do. The operation of performing the decoding may include performing decoding of the information bit based on the value of the constraint bit.

According to one embodiment, the previous decoding result may include a value of a bit corresponding to an order preceding the information bit or a plurality of candidate groups of the bit.

According to one embodiment, when a constraint bit among the at least one constraint bit is a freeze bit, the value of the constraint bit may be determined to be a fixed value. When a constraint bit among the at least one constraint bit is a parity bit, the value of the constraint bit may be determined based on a parity generation method.

According to one embodiment, decoding of the information bit may be performed based on the a-posterior probability of the information bit. The posterior probability may be derived based on the at least one value of the at least one constraint bit.

According to one embodiment, the polar code configuration may include bit index information for identifying each bit index of the at least one information bit and the at least one freeze bit.

According to one embodiment, the polar code configuration may include operation method information for error detection of at least one parity bit of the signal.

According to one embodiment, the polar code configuration may include information related to a parity check matrix for the signal and a generation matrix for the polar code.

According to one embodiment, the SC decoding may be performed based on a belief-propagation technique.

According to embodiments of the present disclosure, a device performed by a device in a communication system or broadcasting system includes: memory; at least one transceiver; And it may include at least one processor. The at least one processor may be configured to receive a signal including bits encoded based on polar codes. The at least one processor may be configured to identify polar code configurations for at least one information bit and at least one frozen bit of the signal. The at least one processor may be configured to decode the signal based on the configuration of the polar code. Decoding of an information bit among the at least one information bit may be performed based on at least one value of at least one constraint bit. The at least one constraint bit may include at least one of a parity bit or a freeze bit corresponding to a later order than the information bit.

According to one embodiment, each bit of the at least one constraint bit may be included in at least one parity bit or the at least one freeze bit of the signal.

According to one embodiment, in order to perform the decoding, the at least one processor determines the value of a constraint bit corresponding to a subsequent information bit based on a previous decoding result of the information bit and candidate values of the information bit. It can be configured to determine . To perform the decoding, the at least one processor may be configured to perform decoding on the information bit based on the value of the constraint bit.

According to one embodiment, the previous decoding result may include a value of a bit corresponding to an order preceding the decoding order of the information bit or a plurality of candidate groups of the bit.

According to one embodiment, when a constraint bit among the at least one constraint bit is a freeze bit, the value of the constraint bit may be determined to be a fixed value. When a constraint bit among the at least one constraint bit is a parity bit, the value of the constraint bit may be determined based on a parity generation method.

According to one embodiment, decoding of the information bit may be performed based on the a-posterior probability of the information bit. The posterior probability may be derived based on the at least one value of the at least one constraint bit.

According to one embodiment, the polar code configuration may include bit index information for identifying each bit index of the at least one information bit and the at least one freeze bit.

According to one embodiment, the polar code configuration may include operation method information for error detection of at least one parity bit of the signal.

According to one embodiment, the polar code configuration may include information related to a parity check matrix for the signal and a generation matrix for the polar code.

According to one embodiment, the SC decoding may be performed based on a belief-propagation technique.

According to embodiments of the present disclosure, an operation method performed by a receiving device in a communication system or broadcasting system includes the process of receiving and processing signals for bits encoded with polar codes to generate input of polar codes, It may include a process of checking the configuration of a polar code for the input, and a process of performing sequential cancellation (SC)-based decoding with the decoder input based on the determined configuration. In the SC-based decoding, decoding and estimation of at least one bit may be performed by considering the configuration (whether or not it is constrained, how it is constrained, etc.) of the at least one bit to be decoded and estimated later. In particular, the sequential decoding of the polar code proposed in this disclosure is performed in consideration of at least one of the so-called future constraint bits, which are constrained bits whose values are not freely determined among the bits to be decoded and estimated later. It is characterized by

The 'bit to be decoded and estimated after the bit to be decoded' refers to a bit that is decoded later in time in the sequential decoding process. According to general notation, if the sequential decoding process is referred to as an ascending order of bit indices, it refers to a bit having a larger index than the bit to be decoded. For example, If the th bit is to be decoded, th, The th bit, etc. are the ‘bits that are later decoded and estimated.’ In decoding the th bit, the future constraint bit has an index Among larger bits, it refers to a frozen bit (or fixed bit) or parity bit whose value is restricted.

As an embodiment of polar code decoding considering the future constraints, SC-check (SCC) decoding substitutes the value that the bit to be decoded can have, 0 or 1, and replaces the value that the bit to be decoded with in front of the bit to be next decoded. It consists of a process of checking or checking whether all or at least one bit of the future constraint bits (with a small index) has a corresponding constraint value.

As an embodiment of SCC decoding, the decoder may substitute the value that the bit to be decoded (hereinafter referred to as target bit), 0 or 1, to the target bit. Afterwards, the decoder can substitute the correspondingly determined bit values into the future constraint bits. The decoder can estimate the value of the target bit by checking or checking whether a contradiction occurs in the configuration of the polar code. In other words, when 0 is assigned to the bit to be decoded and 1 is assigned to the target bit, the decoder can check whether a contradiction occurs in the correspondingly determined bit values of the future constraint bits. The decoder can estimate the target bit value using bits in which no contradiction occurs.

In another embodiment of SCC decoding, the decoder substitutes the values that the bits to be decoded can have, 0 or 1, into the target bits, and then substitutes correspondingly determined bit values for the future constraint bits, and then uses the polar sign. In the configuration of , the value of the target bit can be estimated based on a metric that can be calculated or derived. In other words, the decoder determines the likelihood, posterior probability, etc. of the bit values determined correspondingly of the future constraint bits when the bit to be decoded is substituted as 0 and the bit to be decoded is substituted as 1. By calculating the metric, the estimated value of the target bit can be determined in the direction where the metric value is larger. The specific method of SCC decoding is not limited to the above embodiments. The decoder may not immediately estimate the bit that is the target of decoding, but may substitute 0 and 1 respectively and then make a decision based on at least one of the bits.

As another embodiment of polar code decoding considering the future constraints, belief-propagation SC (BP-SC) decoding is a check node corresponding to at least one future constraint in graph-based polar code decoding. This is a method of performing SC decoding by adding a (check node). As an embodiment of the BP-SC decoding, the decoder can estimate the value of the target bit by performing a corresponding operation on the newly added check node and calculating a metric such as log likelihood ratio (LLR). there is. As another embodiment of the BP-SC decoding, the decoder performs iterative message passing on a graph to which a check node is added, calculates metrics such as LLR for the bit to be decoded, and calculates the value for the target bit. can be estimated.

The SCC decoding and BP-SC decoding can be combined and implemented as a BP-SCC decoding method. In addition, the SCC decoding, BP-SC decoding, and BP-SCC decoding methods are combined with methods such as list and stack that consider two or more candidate groups, so that SCC-list, SCC-stack, SCC-flip, It can be realized in the following ways: BP-SC-list, BP-SC-stack, BP-SC-flip, BP-SCC-list, BP-SCC-stack, BP-SCC-flip. It should be noted that the embodiments of the present disclosure can be implemented in any way that utilizes future constraints in SC-based decoding of polar codes and are not limited in any particular way.

A method performed by a receiving device in a communication system or broadcasting system may include receiving and processing signals for bits encoded with polar codes to generate an input for polar code decoding. The method may include checking the configuration of the polar code for the input. The method may include performing decoding of the polar code based on the configuration of the polar code. The decoding process of the polar code may be characterized in that when decoding at least one bit, it is performed based on the configuration of at least one next bit. According to one embodiment, the bit located next to the at least one may be decoded and estimated later in time or order in sequential decoding of the polar code. According to one embodiment, the value of the bit located next to the at least one may be determined by a pre-determined method or may be a fixed value.

This disclosure relates to a communication system or broadcasting system using polar codes. Specifically, the present disclosure is for decoding polar codes in communication and broadcasting systems. The operating method of the receiving device can generate an input for polar code decoding by receiving and processing signals for bits encoded with polar codes. The operating method may include a process of checking the configuration of the polar sign for the input. The operating method may include a process of performing decoding based on the determined configuration and the input for decoding the polar code. In the decoding process of the polar code, decoding of at least one bit may be performed by considering the configuration of the bit located next to at least one bit. The bit located next to at least one used in the decoding is characterized as a future constraint bit whose value is a constrained bit. The content of the present disclosure is characterized by being distinguished from conventional polar code decoding methods and devices that do not consider future constraints.

Methods according to embodiments described in the claims or specification of the present disclosure may be implemented in the form of hardware, software, or a combination of hardware and software.

When implemented as software, a computer-readable storage medium that stores one or more programs (software modules) may be provided. One or more programs stored in a computer-readable storage medium are configured to be executable by one or more processors in an electronic device (configured for execution). One or more programs include instructions that cause the electronic device to execute methods according to embodiments described in the claims or specification of the present disclosure.

These programs (software modules, software) may include random access memory, non-volatile memory, including flash memory, read only memory (ROM), and electrically erasable programmable ROM. (electrically erasable programmable read only memory, EEPROM), magnetic disc storage device, compact disc-ROM (CD-ROM), digital versatile discs (DVDs), or other types of disk storage. It can be stored in an optical storage device or magnetic cassette. Alternatively, it may be stored in a memory consisting of a combination of some or all of these. Additionally, multiple configuration memories may be included.

In addition, the program may be distributed through a communication network such as the Internet, an intranet, a local area network (LAN), a wide area network (WAN), or a storage area network (SAN), or a combination thereof. It may be stored on an attachable storage device that is accessible. This storage device can be connected to a device performing an embodiment of the present disclosure through an external port. Additionally, a separate storage device on a communications network may be connected to the device performing embodiments of the present disclosure.

In the specific embodiments of the present disclosure described above, elements included in the disclosure are expressed in singular or plural numbers depending on the specific embodiment presented. However, singular or plural expressions are selected to suit the presented situation for convenience of explanation, and the present disclosure is not limited to singular or plural components, and even components expressed in plural may be composed of singular or singular. Even expressed components may be composed of plural elements.

Meanwhile, in the detailed description of the present disclosure, specific embodiments have been described, but of course, various modifications are possible without departing from the scope of the present disclosure.

Claims (20)

In a method performed by a device in a communication system or broadcasting system,
An operation of receiving a signal containing bits encoded based on polar codes;
identifying polar code configurations for at least one information bit and at least one frozen bit of the signal;
An operation of performing decoding of the signal based on the configuration of the polar code,
Decoding of an information bit among the at least one information bit is performed based on at least one value of at least one constraint bit,
The at least one constraint bit includes at least one of a parity bit or a freeze bit corresponding to a later order than the information bit.
The method of claim 1, wherein each bit of the at least one constraint bit is included in at least one parity bit or the at least one freeze bit of the signal.
The method according to claim 1, wherein the operation of performing the decoding includes:
An operation of determining a value of a constraint bit corresponding to a subsequent information bit based on a previous decoding result of the information bit and candidate values of the information bit;
A method comprising performing decoding on the information bit based on the value of the constraint bit.
The method of claim 3, wherein the previous decoding result includes a value of a bit corresponding to an order preceding the information bit or a plurality of candidate groups of the bit,
In claim 1,
If a constraint bit among the at least one constraint bit is a freeze bit, the value of the constraint bit is determined to be a fixed value,
When a constraint bit among the at least one constraint bit is a parity bit, the value of the constraint bit is determined based on a parity generation method.
In claim 1,
Decoding of the information bit is performed based on the a-posterior probability of the information bit,
The method wherein the posterior probability is derived based on the at least one value of the at least one constraint bit.
In claim 1,
The method wherein the polar code configuration includes bit index information for identifying each bit index of the at least one information bit and the at least one freeze bit.
The method of claim 1, wherein the polar code configuration includes operation method information for error detection of at least one parity bit of the signal.
The method of claim 1, wherein the polar code configuration includes information related to a parity check matrix for the signal and a generation matrix for the polar code.
The method of claim 1, wherein the SC decoding is performed based on a belief-propagation technique.
In a device performed by a device in a communication system or broadcasting system,
memory;
at least one transceiver; and
Includes at least one processor,
The at least one processor,
Receive a signal containing bits encoded based on polar codes,
identify polar code configurations for at least one information bit and at least one frozen bit of the signal;
Configured to perform decoding of the signal based on the configuration of the polar code,
Decoding of an information bit among the at least one information bit is performed based on at least one value of at least one constraint bit,
The at least one constraint bit includes at least one of a parity bit or a freeze bit corresponding to a later order than the information bit.
The apparatus of claim 11, wherein each bit of the at least one constraint bit is included in at least one parity bit or the at least one freeze bit of the signal.
The method of claim 11, wherein, to perform the decoding, the at least one processor:
Based on the previous decoding result of the information bit and candidate values of the information bit, determine the value of the constraint bit corresponding to the next information bit,
A device configured to perform decoding on the information bit based on the value of the constraint bit.
The method of claim 13, wherein the previous decoding result includes a value of a bit corresponding to an order preceding the information bit or a plurality of candidate groups of the bit,
In claim 11,
If a constraint bit among the at least one constraint bit is a freeze bit, the value of the constraint bit is determined to be a fixed value,
When a constraint bit among the at least one constraint bit is a parity bit, the value of the constraint bit is determined based on a parity generation method.
In claim 11,
Decoding of the information bit is performed based on the a-posterior probability of the information bit,
The posterior probability is derived based on the at least one value of the at least one constraint bit.
The apparatus of claim 11, wherein the polar code configuration includes bit index information for identifying each bit index of the at least one information bit and the at least one freeze bit.
The device of claim 11, wherein the polar code configuration includes operation method information for error detection of at least one parity bit of the signal.
The device of claim 11, wherein the polar code configuration includes information related to a parity check matrix for the signal and a generation matrix for the polar code.
The apparatus of claim 11, wherein the SC decoding is performed based on a belief-propagation technique.

Images