KR20190013660A - A Method and apparatus for transmitting information using polar code - Google Patents

Abstract

The present invention relates to a method for transmitting information using a polar code, and an apparatus for supporting the same. More particularly, the present invention proposes a method and an apparatus for performing hybrid automatic repeat request (HARQ) retransmissions using multiple cyclic redundancy checks (CRCs) and structured/unstructured polar codes. An information transmission method, in an information transmission of a terminal based on a polar code in wireless communication system, comprises the following steps: generating a first packet comprising a first information bit set, wherein the information is encoded using a polar code; transmitting the first packet, wherein the first packet comprises a plurality of first mini-packets and a plurality of CRCs attached to each of the plurality of first mini-packets; generating a second packet comprising a second information bit set encoded from all or some bits of the information bit set included in the specific first mini-packet among the first information bit set using the polar code, when an error for a specific first mini-packet transmission among the first packets is detected; and transmitting the second packet, wherein the second packet comprises a plurality of CRCs attached to each of the plurality of second mini-packets and the plurality of second mini-packets.

Description

Technical Field [0001] The present invention relates to a method for transmitting information based on polar codes in a wireless communication system,

The following description relates to a wireless communication system, and a method of transmitting and receiving information using a polar code in a wireless communication system and a device supporting the same.

Various devices and technologies such as machine-to-machine (M2M) communication, machine type communication (MTC), and the like requiring a high data transmission amount, such as a smart phone and a tablet PC have. As a result, the amount of data required to be processed in a cellular network is increasing very rapidly. In order to satisfy such a rapidly increasing data processing demand, a carrier aggregation technique, a cognitive radio technique and the like for efficiently using more frequency bands, Multi-antenna technology and multi-base station cooperation technologies are being developed.

As more and more communication devices require greater communication capacity, there is a growing need for enhanced mobile broadband (eMBB) communication over legacy radio access technology (RAT). In addition, massive machine type communication (mMTC) for providing various services anytime and anywhere by connecting a plurality of devices and objects is one of the major issues to be considered in the next generation communication.

In addition, discussions are underway on communication systems to be designed with reliability and latency-sensitive services / UEs in mind. The introduction of the next generation radio access technology is being discussed in consideration of eMBB communication, mMTC, ultra-reliable and low latency communication (URLLC).

(Reference 1) E. Arikan, " Systematic polar coding ", IEEE Communications Letters, vol. 15, pp. 860-862, Aug. 2011.

With the introduction of a new wireless communication technology, not only the number of UEs to which a base station should provide a service in a predetermined resource area increases, but also the amount of data and control information transmitted / received from / . Since the amount of radio resources available for the base station to communicate with the UE (s) is finite, the base station may transmit / receive uplink data and / or uplink / downlink control information to / from the UE (s) A new scheme for efficient reception / transmission is required. In other words, there is a need for a scheme for efficiently utilizing high density nodes or high density user equipments for communication as the density of nodes increases and / or the density of user equipments increases.

Also, as the technology advances, the use of the unused frequency band is being discussed. However, since the newly introduced frequency band is different from the existing frequency band, it is difficult to apply the existing communication technology as it is. Therefore, it is required to introduce a communication technique suitable for a new frequency band used for communication.

Especially, when polar codes are used for HARQ (hybrid automatic repeat request) retransmission, it is necessary to consider coding efficiency and error reduction.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not intended to limit the invention to the particular form disclosed. ≪ / RTI >

The present invention provides a method for transmitting information based on polar codes and a device therefor in a wireless communication system.

As an aspect of the present invention, a method of generating a first packet comprising a first set of information bits, said information being encoded using a polar code; Wherein the first packet comprises a plurality of first mini-packets and a plurality of cyclic redundancy checks (CRC) attached to each of the plurality of first mini-packets; If an error for a specific first mini-packet transmission among the first packets is detected, encoding all or some bits of the information bit set included in the specific first mini-packet among the first information bit set using a polar code Generating a second packet comprising a second set of information bits; And transmitting the second packet, wherein the second packet is comprised of a plurality of CRC (Cyclic Redundancy Check) attached to each of the one or more second mini-packets and the one or more second mini-packets, We propose a transmission method.

In this case, the second information bit set rearranges the information bit set included in the specific first mini-packet based on the reliability of each bit in the information bit set included in the specific first mini-packet, and encodes Bit set.

The second information bit set is a set of bits in which some bits of the information bit set included in the specific first mini packet are encoded and the bits are included in the information bit set included in the specific first mini packet May be selected based on the reliability of the bits.

Also, the first set of information bits and the second set of information bits may be encoded using a systematic polar code.

Also, the first set of information bits may be encoded using a systematic polar code, and the second set of information bits may be encoded using an unstructured polar code.

The second packet may include a parity check bit identical to a parity check bit included in the first packet.

The CRC attached to each of the plurality of first mini packets is attached for error detection to one or more first mini packets, and the CRC attached to each of the one or more second mini packets is one or more Lt; RTI ID = 0.0 > 2 < / RTI >

Also, the second packet may be a set of bits in which only the information bits included in the first mini packet having the smallest number of CRCs attached for error checking among the first information bit set are encoded.

Also, the first packet may be transmitted over a plurality of layers, and the first packet may include a CRC attached for error detection for each of the plurality of layers.

Also, a method for transmitting and receiving signals performed by the base station may include generating a first packet including a first set of information bits, the information being encoded using polar codes; Wherein the first packet comprises a plurality of first mini-packets and a plurality of cyclic redundancy checks (CRC) attached to each of the plurality of first mini-packets; If an error for a specific first mini-packet transmission among the first packets is detected, encoding all or some bits of the information bit set included in the specific first mini-packet among the first information bit set using a polar code Generating a second packet comprising a second set of information bits; And transmitting the second packet, wherein the second packet comprises one or more second mini-packets and a plurality of CRCs attached to each of the one or more second mini-packets.

According to another aspect of the present invention, there is provided a terminal for transmitting information based on a polar code, the apparatus comprising: an encoder configured to encode the information using a polar code; And an RF (Radio Frequency) unit configured to transmit a packet including a set of information bits generated by encoding the information, wherein the encoder encodes the information using a polar code, Generating a first packet; The RF unit transmits the first packet, wherein the first packet is composed of a plurality of first mini packets and a plurality of CRC (Cyclic Redundancy Check) attached to each of the plurality of first mini packets; If an error is detected for a specific first mini-packet transmission among the first packets, the encoder converts all or some bits of the information bit set included in the specific first mini-packet among the first information bit set into a polar code Generating a second packet comprising a second set of information bits encoded using the second set of information bits; And the RF unit transmits the second packet, wherein the second packet consists of one or more second mini-packets and a plurality of CRCs attached to each of the one or more second mini-packets.

A base station for transmitting information based on polar codes in a wireless communication system, the base station comprising: an encoder configured to encode the information using a polar code;

And an RF (Radio Frequency) unit configured to transmit a packet including a set of information bits generated by encoding the information, wherein the encoder encodes the information using a polar code, Generating a first packet;

The RF unit transmits the first packet; Wherein the first packet is composed of a plurality of first mini packets and a plurality of CRC (Cyclic Redundancy Check) attached to each of the plurality of first mini packets; If an error is detected for a specific first mini-packet transmission among the first packets, the encoder converts all or some bits of the information bit set included in the specific first mini-packet among the first information bit set into a polar code Generating a second packet comprising a second set of information bits encoded using the second set of information bits; And the RF unit transmits the second packet, wherein the second packet consists of one or more second mini-packets and a plurality of CRCs attached to each of the one or more second mini-packets.

It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are intended to provide further explanation of the invention as claimed and will become apparent to those skilled in the art from the following detailed description, Can be derived and understood based on the description.

According to the embodiments of the present invention, the following effects are obtained.

According to an embodiment of the present invention, a wireless communication signal can be efficiently transmitted / received. As a result, the overall throughput of the wireless communication system can be increased. Also, in wireless communication systems, signals can be transmitted / received efficiently and at low error rates.

According to the present invention, when polar codes are used for channel coding, the transmission efficiency and the error detection / correction probability can be increased.

The effects obtainable in the embodiments of the present invention are not limited to the effects mentioned above, and other effects not mentioned can be obtained from the description of the embodiments of the present invention described below by those skilled in the art Can be clearly understood and understood. In other words, undesirable effects of implementing the present invention can also be obtained by those skilled in the art from the embodiments of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention. It is to be understood, however, that the technical features of the present invention are not limited to the specific drawings, and the features disclosed in the drawings may be combined with each other to constitute a new embodiment. Reference numerals in the drawings refer to structural elements.
Figure 1 illustrates the processing of a transport block in an LTE / LTE-A system.
2 is a block diagram illustrating performing rate matching by separating a systematic portion and a parity portion of an encoded code block.
Figure 3 shows the internal structure of the circular buffer.
4 is a block diagram for a polar code encoder.
FIG. 5 illustrates the concept of channel combining and channel splitting for channel polarization
FIG. 6 illustrates N-th level channel combining for polar codes.
Figure 7 illustrates the evolution of the decoding paths in the list-L decoding process.
Fig. 8 is shown to illustrate the concept of selecting the location (s) to which the information bit (s) will be allocated in the polar code.
Figure 9 illustrates puncturing and information bit allocation in accordance with the present invention.
10 is a diagram illustrating a HARQ retransmission method using polar coding according to an embodiment of the present invention.
11 is a diagram illustrating information bits in a HARQ retransmission method using polar coding according to an embodiment of the present invention.
12 is a diagram illustrating a method of performing BP polar decoding according to an embodiment of the present invention.
13 is a diagram illustrating a HARQ retransmission method and an error correction method using polar coding according to an embodiment of the present invention.
FIG. 14 is a diagram illustrating a method of using a superposed CRC in a HARQ retransmission method using polar coding according to an exemplary embodiment of the present invention. Referring to FIG.
15 is a diagram illustrating a systematic polar code and an unstructured polar code according to an exemplary embodiment of the present invention.
16 to 20 are diagrams illustrating a HARQ retransmission method combining a systematic polar code and an unstructured polar code according to an exemplary embodiment of the present invention.
21 is a diagram showing a configuration of a terminal and a base station from which the proposed embodiments can be implemented.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following detailed description, together with the accompanying drawings, is intended to illustrate exemplary embodiments of the invention and is not intended to represent the only embodiments in which the invention may be practiced. The following detailed description includes specific details in order to provide a thorough understanding of the present invention. However, those skilled in the art will appreciate that the present invention may be practiced without these specific details.

In some instances, well-known structures and devices may be omitted or may be shown in block diagram form, centering on the core functionality of each structure and device, to avoid obscuring the concepts of the present invention. In the following description, the same components are denoted by the same reference numerals throughout the specification.

The techniques and apparatuses and systems described below can be applied to various wireless multiple access systems. Examples of multiple access systems include a code division multiple access (CDMA) system, a frequency division multiple access (FDMA) system, a time division multiple access (TDMA) system, an orthogonal frequency division multiple access (OFDMA) system, a single carrier frequency division multiple access (MC-FDMA) system, and a multi-carrier frequency division multiple access (MC-FDMA) system. CDMA may be implemented in wireless technologies such as Universal Terrestrial Radio Access (UTRA) or CDMA2000. TDMA may be implemented in wireless technologies such as Global System for Mobile communications (GSM), General Packet Radio Service (GPRS), Enhanced Data Rates for GSM Evolution (EDGE) (i.e., GERAN) OFDMA may be implemented in wireless technologies such as IEEE (Institute of Electrical and Electronics Engineers) 802.11 (WiFi), IEEE 802.16 (WiMAX), IEEE 802-20, and evolved-UTRA (E-UTRA). UTRA is part of Universal Mobile Telecommunication System (UMTS), and 3GPP (Long Term Evolution) is part of E-UMTS using E-UTRA. 3GPP LTE adopts OFDMA in the downlink (DL) and adopts SC-FDMA in the uplink (UL). LTE-Advanced (LTE-A) is an evolutionary form of 3GPP LTE. For convenience of description, it is assumed that the present invention is applied to a 3GPP-based communication system, for example, LTE / LTE-A, NR. However, the technical features of the present invention are not limited thereto. For example, although the following detailed description is based on a mobile communication system based on a mobile communication system corresponding to a 3GPP LTE / LTE-A / NR system, other details unique to 3GPP LTE / LTE-A / And is applicable to any mobile communication system.

In the embodiments of the present invention described below, the expression " assumed " may mean that the channel transmitting entity transmits the channel so as to match the " assumption ". The subject receiving the channel may mean that the channel is received or decoded in a form conforming to the " assumption " on the assumption that the channel is transmitted in conformity with the " assumption ".

In the present invention, the UE may be fixed or mobile and various devices communicating with a base station (BS) to transmit and receive user data and / or various control information. The UE may be a terminal equipment, a mobile station, a mobile terminal, a user terminal, a subscriber station, a wireless device, a personal digital assistant (PDA), a wireless modem ), A handheld device, and the like. Also, in the present invention, a BS is generally a fixed station that communicates with a UE and / or another BS, and exchanges various data and control information by communicating with a UE and another BS. The BS may be referred to as other terms such as Advanced Base Station (ABS), Node-B (NB), Evolved-NodeB (eNB), Base Transceiver System (BTS), Access Point and Processing Server. In particular, the base station of the UTRAN is called Node-B, the base station of E-UTRAN is called eNB, and the base station of the new radio access technology network is called gNB. Hereinafter, for convenience of explanation, the BS is collectively referred to as an eNB.

In the present invention, a node refers to a fixed point that can communicate with a UE to transmit / receive a radio signal. Various types of eNBs can be used as nodes regardless of its name. For example, BS, NB, eNB, pico-cell eNB (PeNB), home eNB (HeNB), relay, repeater and the like can be nodes. Also, the node may not be an eNB. For example, a radio remote head (RRH), a radio remote unit (RRU). RRH, RRU, etc. generally have a lower power level of the eNB. RRH / RRU and RRH / RRU) are generally connected to the eNB as a dedicated line such as an optical cable. Therefore, compared with cooperative communication by eNBs connected by radio lines in general, the RRH / RRU and the eNB Can be performed smoothly. At least one antenna is installed in one node. The antenna may be a physical antenna, an antenna port, a virtual antenna, or an antenna group. A node is also called a point.

In the present invention, a cell refers to a geographical area where one or more nodes provide communication services. Accordingly, in the present invention, communication with a specific cell may mean communicating with an eNB or a node providing a communication service to the specific cell. Also, the downlink / uplink signals of a particular cell are downlink / uplink signals to / from an eNB or a node that provides communication services to the particular cell. A cell providing an uplink / downlink communication service to a UE is called a serving cell. The channel state / quality of a specific cell means the channel state / quality of a channel or a communication link formed between an eNB or a node providing the communication service to the particular cell and the UE. In a 3GPP-based communication system, the UE determines the downlink channel status from a particular node by comparing the CRS (s) transmitted on a cell-specific reference signal (CRS) resource allocated to the particular node with the antenna port / RTI > and / or CSI-RS (s) transmitted on a CSI-RS (Channel State Information Reference Signal) resource.

Meanwhile, a 3GPP-based communication system uses a concept of a cell to manage radio resources, and a cell associated with a radio resource is distinguished from a cell in a geographical area.

The " cell " of a geographical area can be understood as a coverage where a node can provide a service using a carrier, and a " cell " of a radio resource is a frequency range bandwidth, BW). The coverage of the node depends on the downlink coverage where the node can transmit a valid signal and the uplink coverage that can receive a valid signal from the UE depends on the carrier carrying the signal. It is also associated with the coverage of the " cell ". Thus, the term " cell " can sometimes be used to denote the coverage of a service by a node, sometimes to the extent to which a radio resource, and sometimes a signal using the radio resource, can reach a valid strength.

Meanwhile, the 3GPP communication standard uses the concept of a cell to manage radio resources. A cell associated with a radio resource is defined as a combination of DL resources and UL resources, that is, a combination of a DL component carrier (CC) and a UL CC. A cell may be configured to be a DL resource alone, or a combination of DL resources and UL resources. If carrier aggregation is supported, the linkage between the carrier frequency of the DL resource (or DL CC) and the carrier frequency of the UL resource (or UL CC) . For example, a combination of a DL resource and a UL resource can be indicated by linkage of System Information Block Type 2 (SIB2). Here, the carrier frequency means a center frequency of each cell or CC. Hereinafter, a cell operating on a primary frequency will be referred to as a primary cell (Pcell) or a PCC, and a cell operating on a secondary frequency (or SCC) will be referred to as a secondary cell cell, Scell) or SCC. The carrier corresponding to the Pcell in the downlink is called a downlink primary CC (DL PCC), and the carrier corresponding to the Pcell in the uplink is called a UL primary CC (DL PCC). Scell means a cell which can be set after a radio resource control (RRC) connection establishment is made and can be used for providing additional radio resources. Depending on the capabilities of the UE, a Scell may form together with the Pcell a set of serving cells for the UE. The carrier corresponding to the Scell in the downlink is referred to as a DL secondary CC (DL SCC), and the carrier corresponding to the Scell in the uplink is referred to as a UL secondary CC (UL SCC). For UEs that are in the RRC_CONNECTED state but have no carrier aggregation or that do not support carrier aggregation, there is only one serving cell that is only configured with the Pcell.

The 3GPP-based communication standard includes downlink physical channels corresponding to resource elements carrying information originating from an upper layer and downlink physical channels used by the physical layer but corresponding to resource elements not carrying information originated from an upper layer Physical signals are defined. For example, a Physical Downlink Shared Channel (PDSCH), a Physical Broadcast Channel (PBCH), a Physical Multicast Channel (PMCH), a Physical Control Format Indicator Channel a physical downlink control channel (PDCCH), and a physical hybrid ARQ indicator channel (PHICH) are defined as downlink physical channels, and a reference signal and a synchronization signal Are defined as downlink physical signals. A reference signal (RS), also referred to as a pilot, refers to a signal of a particular predetermined waveform that is known to the UE and the eNB, for example a cell specific RS, a UE- A specific RS (UE-specific RS, UE-RS), a positioning RS (PRS) and channel state information RS (CSI-RS) are defined as downlink reference signals. The 3GPP-based communication standard includes uplink physical channels corresponding to resource elements carrying information originating from an upper layer and uplink channels corresponding to resource elements used by the physical layer but not carrying information originated from an upper layer The physical signals are defined. For example, a physical uplink shared channel (PUSCH), a physical uplink control channel (PUCCH), and a physical random access channel (PRACH) And a demodulation reference signal (DMRS) for the uplink control / data signal and a sounding reference signal (SRS) used for the uplink channel measurement are defined.

In the present invention, the Physical Downlink Control Channel (PDCCH) / Physical Control Format Indicator CHannel / Physical Uplink Shared CHannel (PHICH) / Physical Downlink Shared CHannel (PDSCH) A Physical Uplink Control CHannel (PUCCH), a Physical Uplink Control Channel (PUSCH), a Physical Uplink Control Channel (PUSCH), and a Physical Uplink Control Channel (PUSCH) (Uplink Shared CHannel) / PRACH (Physical Random Access CHannel) refers to a set of time-frequency resources or a set of resource elements each carrying Uplink Control Information (UCI) / uplink data / random access signals. / PCFICH / PHICH / PDSCH / PUCCH / PUSCH / PRACH RE or a time-frequency resource or resource element (RE) assigned to or belonging to the PDCCH / Hereinafter, the expression that the user equipment transmits a PUCCH / PUSCH / PRACH is referred to as a PUCCH / PUCCH / PRACH or a PUCCH / PUCCH / PRACH through an uplink control information / uplink The expression that the eNB transmits PDCCH / PCFICH / PHICH / PDSCH is used to transmit downlink data / control information on the PDCCH / PCFICH / PHICH / PDSCH, Is used in the same sense.

The terms and techniques used in the present invention are not specifically described in the 3GPP LTE / LTE-A standard documents such as 3GPP TS 36.211, 3GPP TS 36.212, 3GPP TS 36.213, 3GPP TS 36.321 and 3GPP TS 36.331 and the like and 3GPP NR standard documents such as 3GPP TS 38.211, 3GPP TS 38.212, 3GPP TS 38.213, 3GPP TS 38.214, 3GPP TS 38.300, 3GPP TS 38.331, etc. In addition, the principle of encoding and decoding using polar codes and polar codes is described in 'E. Arikan, "Channel Polarization: A Method for Constructing Capacity-Achieving Codes for Symmetric Binary-Input Memoryless Channels," in IEEE Transactions on Information Theory, vol. 55, no. 7, pp. 3051-3073, July 2009).

As more and more communication devices require greater communication capacity, there is a need for improved mobile broadband communication over existing radio access technology (RAT). Also, massive MTC, which provides various services by connecting multiple devices and objects, is one of the major issues to be considered in the next generation communication. In addition, a communication system design considering a service / UE sensitive to reliability and latency is being discussed. The introduction of next-generation RAT, which takes into account advanced mobile broadband communications, massive MTC, URLLC (Ultra-Reliable and Low Latency Communication), is being discussed. Currently, 3GPP is conducting studies on next generation mobile communication systems after EPC. For convenience, the present invention is referred to as a new RAT (new RAT, NR) or 5G RAT.

NR communication systems are required to support significantly better performance than existing fourth generation (4G) systems in terms of data rate, capacity, latency, energy consumption and cost. Thus, NR systems need to make considerable progress in the areas of bandwidth, spectral, energy, signaling efficiency, and cost per bit. NR needs to utilize an efficient waveform to meet this demand.

Figure 1 illustrates the processing of a transport block in an LTE / LTE-A system.

In order to correct the error experienced by the channel at the receiving end, the information sent from the transmitting end is encoded using a forward error correction code and then transmitted. The receiver demodulates the received signal and decodes the error correction code to recover the transmission information. In this decryption process, the error on the received signal caused by the channel is corrected.

The data reaches the coding block with the behavior of a maximum of two transport blocks every TTI for each DL / UL cell. The following coding steps can be applied for each transport block of the DL / UL cell:

Add CRC to transport block;

Code block segmentation and code block CRC attachment;

Channel coding;

Rate matching;

Code block concatenation.

Various error correction codes can be used, but Turbo codes are mainly used in existing LTE / LTE-A systems. The turbo code consists of a recursive systematic convolution encoder and an interleaver. One of the interleavers is quadratic polynomial permutation (QPP) to facilitate parallel decoding in actual implementation of turbo codes. Such a QPP interleaver is known to maintain good performance only for a specific data block size. The performance of the turbo code is known to be better as the size of the data block increases. In actual communication systems, encoding is performed by dividing a data block of a predetermined size or larger into a plurality of small data blocks for practical implementation convenience. The divided small data block is called a code block. The code block generally has the same size, but because of the size limitation of the QPP interleaver, one of the code blocks may have different sizes. An error correction coding process is performed in units of code blocks of a predetermined interleaver size, and then interleaving is performed to reduce the influence of a burst error occurring in transmission on a wireless channel. Then, it is mapped to actual radio resources and transmitted. Since the amount of radio resources used in the actual transmission is constant, rate matching must be performed on the encoded code block in order to match it. In general, rate matching is done by puncturing or repetition. For example, if the amount of radio resources, that is, the number of transmission bits that can be transmitted by the corresponding radio resource is M, and the number of output bits of the encoder is N, Rate matching is performed to adjust the length of the coded bit sequence to match with M. If M > N, all or some of the bits of the coded bit sequence are repeated so that the length of the rate matched sequence is equal to M. [ If M < N, some of the bits of the coded bit sequence are punctured such that the length of the rate matched sequence is equal to M, and the punctured bits are excluded from transmission.

That is, in the LTE / LTE-A system, data to be transmitted is encoded using channel coding having a specific code rate (for example, 1/3), and then a code rate of data to be transmitted through a rate matching process, . When a turbo code is used as a channel code in LTE / LTE-A, a process of channel coding and rate matching of each code block in the transport channel processing process as shown in FIG. 1 is shown in FIG.

2 is a block diagram illustrating performing rate matching by separating a systematic portion and a parity portion of an encoded code block.

As shown in FIG. 2, the mother code rate of the LTE / LTE turbo encoder is 1/3. In order to obtain different code rates, repetition or puncturing must be performed, if necessary, and these are done by the rate matching module. The rate matching module comprises three so-called sub-block interleavers for the three output streams of the turbo encoder and bit selection and pruning (b), realized by a circular buffer pruning part. The sub-block interleaver is based on a classical row-column interleaver with 32 rows and a length-32 intra-column permutation. The bits of each of the three streams are written in a matrix with 32 columns by row-by-row (the number of rows depends on the stream size). Dummy bits are padded in front of each stream to completely fill the matrix. After the thermal permutation, the bits are read from the matrix in column-by-column.

Figure 3 shows the internal structure of the circular buffer.

The circular buffer is the most important part of the rate matching module, which allows puncturing and repetition of the mother code. Referring to FIG. 2, the interleaved systematic bits are written in the cyclic buffer in sequence, with the first bit of the interleaved systematic bitstreams at the beginning of the cyclic buffer. The interleaved and interleaved parity bit streams are sequentially written to the circular buffer with the first bit of the stream after the last bit of the interleaved systematic bit stream. The coded bits are read serially from a certain starting point specified by redundancy version (RV) points in the cyclic buffer (depending on the code rate). If the end of the cyclic buffer is reached and more coded bits are needed for transmission (e.g., for code rates less than 1/3), the transmitting device wraparounds and continues at the beginning of the cyclic buffer continue).

HARQ indicating hybrid ARQ is an error correction mechanism based on retransmission of packets detected to be erroneous. The transmitted packet arrives at the receiving device after a certain propagation delay. The receiving apparatus produces an ACK in the case of an error-free transmission and produces a NACK when an error is detected. The ACK / NACK is produced after some processing time and is sent to the transmitting device and arrives at the transmitting device after the propagation delay. NACK In this case, after some processing delay at the transmitting device, the desired packet will be resent. The bits read from the circular buffer and sent in each retransmission are different and depend on the location of the RV. There are four RVs (0, 1, 2, 3) that define the location of the start point at which bits are read from the circular buffer. Referring to FIG. 3, as the number of retransmissions progresses, the RV increases, so fewer systematic bits and more parity bits are read from the circular buffer for retransmission.

NR now offers better speed and coverage than 4G, operates in high frequency bands, provides speeds up to 1 Gb / s for dozens of connections, or up to tens of Mb / s for tens of thousands of connections Is required. In order to meet the requirements of such NR systems, the introduction of advanced coding schemes rather than existing coding schemes is being discussed. Because data communication occurs in an incomplete channel environment, channel coding plays an important role in achieving higher data rates for fast and fault-free communication. The selected channel code should have excellent block error ratio (BLER) performance at a certain range of block lengths and code rates. Here, BLER is defined as a ratio of the number of erroneous receiving blocks to the total number of transmitted blocks. In NR, low computational complexity, low latency, low cost, and greater flexibility are required in the coding scheme. Furthermore, reduced energy per bit and improved area efficiency are required to support higher data rates. eMBB, Massive IoT, URLLC, etc. are considered to be examples of use of NR networks. eMBB covers Internet access with high data rates to enable rich media applications, cloud storage and applications, and augmented reality for entertainment. Massive IoT applications include dense sensor networks for smart homes / buildings, remote health monitoring, and logistics tracking. URLLC covers critical applications requiring ultra-high reliability and low latency, such as industrial automation, unmanned vehicles, remote surgery, and smart grids.

While many coding schemes are available with high capacity performance at large block lengths, many of them do not consistently show good performance over a wide range of block lengths and code rates. However, turbo codes, low density parity check (LDPC) codes and polar codes show promising BLER performance over a wide range of coding rates and code lengths, Use is being considered. There is a need for a coding scheme that provides stronger channel coding efficiency than turbo codes as the demand for various cases such as eMBB, massive IoT and URLLC increases. In addition, an increase in the maximum number of subscribers that the channel can currently accommodate, i.e., an increase in capacity, is also required.

The polar code is a code that provides a new framework for solving the problems of existing channel codes, and was invented by Arikan of Bikent University (see E. Arikan, "Channel Polarization: A Method for Constructing Capacity-Achieving Codes for Symmetric Binary-Input Memoryless Channels, " in IEEE Transactions on Information Theory, vol. 55, No. 7, pp. 3051-3073, July 2009). The polar code is the first mathematically proven, capacity-achieving code with low encoding and decoding complexity. The polar code outperforms the turbo code in large block lengths without any error flow. Hereinafter, channel coding using a polar code is referred to as polar coding.

Polar codes are known as numeric codes that achieve channel capacity in a given binary discrete memoryless channel. This can only be done if the block size is large enough. That is, the polar code is a code capable of achieving the channel capacity when the size N of the code is infinitely increased. The polar codes are less complex in encoding and decoding and can be successfully decoded. Polar codes are a type of linear block error correcting code, and a number of recursive concatenations are the basic building blocks for polar codes and are the basis for code construction. A physical conversion of the channel that translates the physical channels into the virtual channels takes place, and this conversion is based on a number of concatenated regressions. When multiple channels are multiplied and accumulated, most of the channels are getting better or worse, and the idea behind the polar code is to use good channels. For example, sending data at rate 1 over good channels and sending it at rate 0 over bad channels. That is, through channel polarization, the channels enter a polarized state from a normal state.

4 is a block diagram for a polar code encoder.

FIG. 4A shows a base module of a polar code, in particular, a first level channel combining for polar coding. In FIG. 4 (a), W2 means the entire equivalent channel obtained by combining two W-DMC (W-DMC) channels. Where u1 and u2 are binary-input source bits, and y1 and y2 are output coded bits. Channel combining is a process of concatenating B-DMC channels in parallel.

4B shows a basic matrix F for the basic module. The binary-input source bits u1 and u2 of the basic matrix F and the corresponding outputs x1 and x2 have the following relationship.

The channel W2 can achieve the symmetry capacity I (W) which is the highest rate. The symmetric capacity in B-DMC W is an important parameter, the symmetric capacity is used for the rate measurement, and the highest rate at which reliable communication can occur across channel W. The B-DMC can be defined as follows.

It is possible to synthesize or create a second set of N binary input channels from N independent copies of a given B-DMC W, i): 1? i? N}. As N increases, some of the channels tend to be channels with capacities close to 1, while others tend to be channels with capacities close to zero. This is called channel polarization. In other words, channel polarization is a process of generating a second set of N channels {WN (i): 1? I? N} using N independent copies of a given B-DMC W, It means that all symmetric capacity terms {I (WN (i)) tend to be either 0 or 1, except for the fraction vanishing indexes i. In other words, the concept behind channel polarization in polar cords is that N copies of an additive white Gaussian noise channel (i.e., N transmissions) of a symmetric capacity channel of I (W) ) Into 1 or extreme channels of close capacity. The I (W) fraction of the N channels will be perfect channels and the 1-I (W) portion will be completely noise channels. The information bits are then only sent on good channels, and inputs to other channels are frozen to 1 or 0. The amount of channel polarization increases with the block length. Channel polarization is composed of two phases: channel combining phase and channel splitting phase.

FIG. 5 illustrates the concept of channel combining and channel splitting for channel polarization. 5, if N copies of the original channel W are appropriately combined to create a vector channel Wvec and then split into new polarized channels, in case of a sufficiently large N, the polarized new channels are respectively allocated to channels Capacities C (W) = 1 and C (W) = 0 are distinguished. In this case, since the bit passing through the channel having the channel capacity C (W) = 1 can be transmitted without error, the information bit is transmitted to the channel having the channel capacity C (W) = 1 and the channel capacity C (W) = 1 It is better to send frozen bits which are meaningless bits.

Referring to Figure 5, copies of a given B-DMC W may be combined in a regressive manner to output the vector channel Wvec given by WN: XN - &gt; YN. Where N = 2n and n is an integer greater than or equal to zero. Recursion always starts at the zeroth level, and W1 = W. n = 1 means the regression of the first level that two independent copies of W1 combine together. When the two copies are combined, a channel W2: X2 - &gt; Y2 is obtained. The transitional probability of this new channel W2 can be expressed by the following equation.

Once the channel W2 is obtained, a single copy of the channel W4 may be obtained by combining the two copies of W2. This regression can be represented by W4: X4 - &gt; Y4 with the following transition probability.

5, GN is a generator matrix of size N. G2 corresponds to the basic matrix F shown in Fig. 4 (b). G4 can be expressed by the following matrix.

Where ⓧ is the Kronecker product, Aⓧn = AⓧAⓧ (n-1) for all n ≥1, and Aⓧ0 = 1.

The relationship between the input uN1 and the output xN1 to GN in Fig. 5 (b) can be expressed as xN1 = uN1GN. Where xN1 = {x1, ..., xN}, uN1 = {u1, ..., uN}.

When combining N B-DMCs, each B-DMC can be represented in a recursive form. That is, GN can be expressed by the following equation.

Here, N = 2n, n? 1, Fⓧn = FⓧFⓧ (n-1), and Fⓧ0 = 1. BN is a permutation matrix known as bit-reversal and can be compiled recursively as BN = RN (I2ⓧBN / 2). I2 is a two-dimensional identity matrix, and this recursion is initialized to B2 = I2. RN is a bit-reversal interleaver and is used to map the inputs sN1 = {s1, ..., sN} to the outputs xN1 = {s1, s3, ..., sN-1, s2, ..., sN} do. RN is first base-2 extension. The bit-reversal interleaver may not be included in the transmission stage. The relationship of Equation (6) is shown in Fig.

FIG. 6 illustrates N-th level channel combining for polar codes.

The process of defining an equivalent channel for a particular input after combining N B-DMC W is called channel splitting. The channel splitting can be expressed by the channel transition probability as the following equation.

Channel polarization has the following characteristics:

Conservation: C (W-) + C (W +) = 2C (W),

Extremization: C (W-)? C (W)? C (W +).

When channel combining and channel splitting are performed, the following theorem can be obtained.

Theorem: For any B-DMC W, channels {WN (i)} are polarized in the following sense. For any fixed δ ∈ {0, 1}, the indices i (WN (i)) ∈ (1 - δ, 1), as N goes infinitely through the power of 2, The fraction of ∈ {1, ..., N} goes to I (W) and the fraction I (WN (i)) ∈ [0, δ) goes to 1-I (W). Therefore, if N → ∞, the channels are completely noise or are polarized freely to noise, and these channels can be known exactly at the transmitting end. Thus, the bad channels can be fixed and the uncommitted bits can be transmitted on good channels.

That is, when the polar code size N becomes infinite, the channel becomes a noisy or noisy channel with respect to a specific input bit. This means that the capacity of the equivalent channel for a particular input bit is divided by 0 or I (W).

An input of a polar encoder is divided into a bit channel to which information data is mapped and a bit channel to which information data is mapped. As described above, according to the theory of polar codes, input bit channels can be divided into noise-free channels and noise channels as the codewords of polar codes go to infinity. Therefore, when information is allocated to a bit channel with no noise, the channel capacity can be obtained. However, since the code word of infinite length can not be constructed in practice, the reliability of the input bit channel is calculated and data bits are allocated in that order. In the present invention, a bit channel to which data bits are allocated is called a good bit channel. A good bit channel is an input bit channel to which data bits are mapped. A bit channel to which data is not mapped is called a frozen bit channel, and a frozen bit channel is encoded by inputting a known value (e.g., 0). Any value known to the transmitting and receiving end may be mapped to the frozen bit channel. When performing puncturing or repetition, information about a good bit channel can be utilized. For example, the position of a codeword bit (i.e., an output bit) corresponding to an input bit position that is not assigned to an information bit may be punctured.

The decoding scheme of the polar code is a successive cancellation (SC) decoding scheme. The SC decoding scheme calculates a channel transition probability and calculates a likelihood ratio (LLR) of the input bits. At this time, the channel transition probability can be calculated in a recursive manner by using the characteristic that the channel combining and the channel splitting are made in a recursive manner. Therefore, finally, the LLR value can also be calculated in a recursive form. First, the channel transition probability WN (i) (y1N, u1i-1 | u1) for the input bit ui can be obtained as follows. u1i may be divided into an odd index and an even index and expressed as u1, oi, u1, and ei. The channel transition probability can be expressed by the following equations.

The polar decoder retrieves information and generates an estimate uN1 of uN1 with known values (e.g., received bits, frozen bits, etc.) in the polar code. The LLR is defined as follows.

The LLR can be calculated recursively as follows.

The recursive computation of LLRs is traced back to a code length of 1 with LLR L (1) 1 (yi) = W (yi | 0) / W (yi | L (1) 1 (yi) is the soft information observed from the channel.

The complexity of polar encoders and SC decoders depends on the length N of polar codes, which is known to have a complexity of O (NlogN). Assuming K input bits in a polar code of length N, the coding rate is N / K. If the generator matrix of the polar encoder of the data payload size N is denoted as GN, the encoded bit can be expressed as xN1 = uN1GN, K bits of uN1 correspond to payload bits, And a row index of GN corresponding to the remaining NK bits is assumed to be F. [ The minimum distance of such a polar code can be given as dmin (C) = mini? I2wt (i). Where wt (i) is the number of 1s in the binary expansions of i and i = 0, 1, ..., N-1.

SC List (SCL) decoding is an extension of the basic SC decoder. In this kind of decoder, L decoding paths are considered simultaneously at each stage of decoding. Where L is an integer. In other words, in the case of polar codes, the list-L decoding algorithm is an algorithm that simultaneously tracks L paths in the decoding process.

Figure 7 illustrates the evolution of decoding paths in the list-L decoding process. For convenience of explanation, it is assumed that the number of bits to be determined is n, and all the bits are not frozen. If the list size L = 4, each level has at most 4 nodes with paths continuing downward. The discontinuous paths are indicated by dashed lines in Fig. Referring to FIG. 7, a description will be given of a process in which decoding paths evolve in the list-L decoding. i) List-L decoding begins, and the first unfrozen bit may be 0 or 1. ii) List-L decoding continues. The second non-frozen bits may be 0 or 1. Since the number of paths is not more than L = 4, pruning is not necessary yet. iii) Considering all the options for the first bit (i.e. the bits of the first level), the second bit (i.e. the bits of the second level) and the third bit (i.e. the bits of the third level) Resulting in a decoding path, and since L = 4, there are too many eight decoding paths. iv) prune the eight decoding paths into L = 4 promising paths. v) continue the four active paths by considering the two options of the fourth non-frozen bit. In this case, the number of paths is doubled to 8, and since L = 4, the number of paths is too large. vi) Prune again with L = 4 best paths. In the example of FIG. 7, four candidate code words 0100, 0110, 0111 and 1111 are obtained, and one of them is determined as the code word most similar to the original code word. As in the normal decoding process, for example, in the process of determining the pruning process or the final codeword, the path with the largest sum of absolute LLR values may be selected as the survival path. If there is a CRC, the survival path may be selected via the CRC.

On the other hand, CRC aided SCL decoding improves the performance of polar codes by SCL decoding using CRC. CRC is the most widely used technique for error detection and error correction in information theory and coding. For example, if the input block to the error correction encoder is K bits, the length of the information bits is k, and the length of the CRC sequence is m bits, K = k + m. The CRC bits are part of the source bits for the error correction code, and if the size of the channel code used for encoding is N, the code rate R is defined as R = K / N. The CRC auxiliary SCL decoding aims at detecting an error-free path while checking a cyclic redundancy check (CRC) code for each path in the receiving apparatus. The SCL decoder outputs the candidate sequences to a CRC detector, which feeds back the check results to assist in codeword determination.

SCL decoding or CRC secondary SCL decoding is more complicated than SC algorithm, but has an advantage of excellent decoding performance. For more information about the list of polar codes-X decoding algorithm, see 'I. Tal and A. Vardy, &quot; List decoding of polar codes, &quot; in Proc. IEEE Int. Symp. Inf. Theory, pp. 1-5, Jul. 2011 '.

Polar code has the disadvantage that code design is independent of the channel, so there is no versatile of mobile fading channels, and relatively recently introduced code is not yet matured and is applied only in a limited manner. That is, the polar coding proposed so far has not been defined to be applied to a wireless communication system as it is. Accordingly, the present invention proposes a polar coding method suitable for a wireless communication system.

Fig. 8 is shown to illustrate the concept of selecting the location (s) to which the information bit (s) will be allocated in the polar code.

In Fig. 8, it is assumed that the size N of the mother code is 8, that is, the polar code size is N = 8, and the code rate is 1/2.

In Fig. 8, C (Wi) is the capacity of the channel Wi, which corresponds to the reliability of the channels that the input bits of the polar code will experience. Assuming that the channel capacities corresponding to the input bit positions of the polar code are as shown in FIG. 8, the reliability of the input bit positions can be ranked as shown in FIG. In this case, in order to transmit data at a code rate of 1/2, the transmitting apparatus transmits the four bits constituting the data to four input bit positions (that is, 8, U6, U7 and U4 among the input bit positions U1 to U8 of 8), and the rest of the input bit positions are frozen. The generator matrix G8 corresponding to the polar code of FIG. 8 is as follows. The generator matrix G8 may be obtained based on Equation (6).

The input bit positions indicated by U1 to U8 in FIG. 8 correspond one-to-one to the rows from the lowest row to the highest row of G8. Referring to FIG. 8, it can be seen that the input bit corresponding to U8 affects all the output coded bits. On the other hand, it can be seen that the input bit corresponding to U1 only affects Y1 among the output coded bits. Referring to Equation 12, a row that causes the input bits to appear in all output bits when the binary-input source bits U1 to U8 and G8 are multiplied is the least significant row of all rows of the G8, Is the row [1, 1, 1, 1, 1, 1, 1, 1]. On the other hand, a row that causes the corresponding binary-input source bit to appear only in one output bit is a row of one of the rows of G8, i.e., [1, 0, 0, 0, 0, 0, 0, 0]. Likewise, a row having a row weight of 2 may reflect input bits corresponding to the row in two output bits. Referring to FIG. 8 and FIG. 12, U1 to U8 correspond to the rows of G8 in a one-to-one correspondence, and bit indices for identifying the input positions to the input positions of U1 to U8, have.

In the polar code, it can be assumed that bit indices are sequentially allocated from the bit index 0 to N-1 starting from the top row having the smallest row weight for N input bits to GN. For example, referring to FIG. 8, a bit index 0 is assigned to the input position of U1, that is, the first row of G8, and a bit index 7 is assigned to the input position of U8, that is, the last row of G8. However, since the bit indices are used to indicate the input positions of polar codes, they can be assigned differently. For example, bit indices 0 to N-1 may be allocated starting from the lowest row with the largest row weight.

In the case of the output bit index, as illustrated in FIGS. 8 and 12, bit indexes 0 to N-1 from the first column with the largest column weight to the last column with the smallest column weight, It can be assumed that the indexes 1 to N are assigned.

In polar code, setting information bits and frozen bits is one of the most important factors in the configuration and performance of polar codes. That is, determining the rank of the input bit positions is an important factor in the performance and configuration of the polar code. For polar codes, the bit indices can identify the input or output positions of polar codes. For a polar code, a sequence obtained by ascending the reliability of the bit positions in ascending order or descending order is called a bit index sequence. That is, the bit index sequence represents the reliability of input or output bit positions of polar codes in ascending or descending order. The transmitting apparatus inputs information bits to the input bits with high reliability based on the input bit index sequence and performs encoding using polar codes. The receiving apparatus performs the encoding using the same or a corresponding input bit index sequence, Input positions or input positions to which frozen bits are assigned. That is, the receiving apparatus can perform polar decoding using the same or a corresponding input bit sequence and the corresponding polar code as the input bit index sequence used by the transmitting apparatus. For polar codes, it can be assumed that the input bit sequence is predetermined to allow the information bit (s) to be assigned to the input bit position (s) with high reliability.

Figure 9 illustrates puncturing and information bit allocation for polar codes. In Fig. 9, F represents the frozen bit, D represents the information bit, and 0 represents the skipping bit.

The information bit may be changed to the frozen bit depending on the index or position of the punctured bit among the coded bits. For example, if the output coded bits for the N = 8 mother code are to be punctured in the order of Y8, Y7, Y6, Y4, Y5, Y3, Y2, Y1, , Y8, Y7, Y6 and Y4 are punctured and U8, U7, U6 and U4 connected only to Y8, Y7, Y6 and Y4 are frozen to zero and these input bits are not transmitted, as illustrated in Fig. An input bit changed to a frozen bit by puncturing a coded bit is called a skipping bit or a shortening bit, and the corresponding input position is referred to as a skipping position or a shortening position. Shortening is a rate matching method in which the size of the input information (i.e., the size of the information block) is maintained, and a known bit is inserted into an input bit position connected to a desired output bit position. The generator matrix GN can be shortened from the input corresponding to the column with a column weight of 1, and the row and column with the column weight 1 can be eliminated and the input corresponding to the column with the column weight 1 again in the remaining matrix can be shortened to the next. The information bits that should have been allocated to the information bit positions to prevent all of the information bits from being punctured can be reallocated in the order of high reliability in the frozen bit position set.

In the case of polar codes, decoding is generally performed in the following order:

The least reliable bit (s) are recovered first. Although it depends on the structure of the decoder, since the reliability of the input index (hereinafter referred to as the encoder input index) is low, the decoding is generally performed sequentially from the encoder input index is small.

If there is known bit information for the recovered bit, the known bit is used together with the recovered bit, or the unknown bit is directly used for a specific input bit position, Restores the bit. The information bits may be source information bits (e.g., bits of a transport block), or may be CRC bits.

1. Polar code using multiple CRC

Polar code based HARQ retransmission using multiple CRC

Hereinafter, a method of using polar codes using multiple CRCs for data retransmission based on hybrid automatic repeat request (HARQ) is proposed. In this method, as shown in FIG. 10A, several CRCs are first inserted between information bits and then polar encoded. In this case, each information block represents a mini-packet, and a plurality of mini-packets are gathered to form one packet. The total length of the packet is k bits, and the entire packet is polar-encoded . The encoded polar codeword has a length of n bits, which is transmitted to the channel. At the receiving end, first polar decoding is performed, and thereafter, a CRC check (CRC check) is performed. 10B, when a CRC check detects that an error has occurred in a specific mini-packet, the receiving end sends a signal to the transmitting end so that only the mini-packet is retransmitted, and the transmitting end retransmits only the specific mini-packet in which the error occurred. At this time, the mini-packet to be retransmitted can be transmitted as it is without any information bit in which polar coding is not applied.

In the present invention, various new ideas are presented to improve the performance of the existing HARQ scheme based on polar codes using multiple CRC. First, in the proposed method, polar coding is used also in retransmission as shown in FIG. 11A. In this case, the horizontal axis represents the index of the bit channel of polar coding, and the vertical axis represents the mutual information corresponding to each bit channel. In the example of Fig. 11, it is assumed that a code having a length of 16 and a code rate of 1/2 is used for the initial transmission. Since the code rate is 1/2, 8 information bits must be transmitted, and these 8 information bits are transmitted on the 8 bit channels having the highest mutual information amount. The eight information bits are divided into two groups of four, and each group is referred to as a mini-packet in the present invention. A CRC is appended to each mini-packet to check whether an error has occurred in the mini-packet. In Fig. 11A, it is assumed that an error has occurred in the first mini-packet. In the conventional scheme, four information bits are retransmitted without using any coding in retransmission, but in the proposed scheme, a polar code of length 8 and code rate 1/2 is used in retransmission. When the retransmission is performed in this manner, the receiver performs polar decoding using two packets at the same time. First, when SIC (successive interference cancellation) or SILC (successive interference list cancellation) scheme is used, it is possible to perform parallel decoding using two packets for transmission. If polar decoding is performed using belief propagation (BP), the LLR (log likelihood ratio) is obtained for each packet, the obtained LLR value is added to each information bit, LLR values are used to determine each information bit.

11B shows a method of optimally aligning the order of information bits in retransmitted packets. In FIG. 11B, the left graph of the two graphs is the first packet to be transmitted, and the first mini packet includes four information bits (A, B, C, and D). At this time, considering the characteristic of polar coding, the four information bits are not transmitted to the channel with the same reliability. That is, since different information bits have different degrees of polarization, mutual information amounts of the bit channels through which the information bits are transmitted are different from each other. In Fig. 11B, bit D has the highest reliability and B has the lowest lightness. In order to compensate for this unbalance of reliability, the information bits are rearranged in the reverse order in the packets to be retransmitted. That is, the information bit B is transmitted with the most reliable polar coding bit channel, and the information bit D is transmitted with the least reliable bit channel. When the polarity coding is performed at the receiving end, the LLR values obtained from the two packets are added to each information bit to finalize the information bit decision, thereby canceling the reliability imbalance between the other information bits.

FIG. 11C shows a proposed HARQ scheme for retransmissions many times. In order to correct the error in the originally transmitted packet through the minimum number of retransmissions, the position of the errored information bit should be estimated as soon as possible in the most efficient manner and the packet must be retransmitted. To this end, the present invention uses the concept of binary search or interval halving. As shown in the example of FIG. 11C, if it is recognized that an error has occurred in one mini-packet, the information bit of the mini-packet is divided again into two or several mini-packets in the second transmission (i.e., the first retransmission) Adds a CRC to the end of the packet. In the receiving end, the CRCs of the second transmitted packet are checked to know the error occurrence position more accurately. If the error is not corrected through the first retransmission, a third transmission (i.e., a second retransmission) is performed. In this case, the first retransmission transmits an information bit corresponding to the error-minimizing packet. It is divided into groups or groups and the CRC is added to the end of each group (mini-packet). This is a concept of binary search. Using this method, the position of the information bit where an error occurs is known through minimum retransmission, and error recovery is possible as fast as possible.

11D proposes a more efficient retransmission scheme. It is assumed that an error occurs in the second mini packet in the first transmission (i.e., when the CRC ₂ ⁽¹⁾ check fails). As shown in FIG. 11D, six information bits all belong to the second mini-packet. One retransmission scheme is to retransmit all these six information bits. However, as described above, the reliability of these six information bits is all different. In the example given in FIG. 11D, the reliability of the information bit F is the highest and the reliability of the information bit C is the lowest. Therefore, instead of retransmitting all six information bits at the time of retransmission, only a few least reliable bits of information can be retransmitted. In the example of Fig. 11D, only the three least significant bits (B, C, D) are retransmitted. When the retransmission is performed in this manner, the channel capacity required for retransmission can be reduced, so that more efficient retransmission can be performed. If the same channel capacity is used, the code rate can be reduced, so that the reliability of the three bits to be retransmitted is further improved.

FIG. 11E shows a comprehensive example of HARQ retransmission performed using the above-described scheme.

Polar code BP decoding method using multiple CRC

In the past, research has been conducted to reduce the complexity and memory capacity required for polar decoding by using multiple CRCs for SILC. Also, in the past, methods of early stopping using one CRC for polar BP decoding have been studied. However, a method of using multiple CRCs for polar BP decoding has not been proposed so far.

In the present invention, as shown in FIG. 12, a method of using multiple CRCs for Polar BP multicoding is proposed. The proposed method will be described in more detail. When using the BP scheme with polar decoding, each CRC is checked whenever iteration is performed. If the CRC check is passed, the LLR value of the corresponding information bit is adjusted to positive infinity (information bit is 1) or music infinity (information bit is 0). In particular, given the undetected error probability (UEP) problem of the CRC, the CRC check is not performed from the first iteration, iteration) value (I _min ).

In FIG. 12, information bits are divided into a plurality of groups, and a CRC is applied to each group. There are various methods of dividing information bits. First, the entire information bits are arranged in descending order, and a predetermined number of information bits are sequentially allocated to each group.

Alternatively, the entire information bits may be arranged according to the reliability of each information bit (i.e., each information bit is listed according to the reliability of the bit channel to which each information bit is transmitted), and a predetermined number of information bits are sequentially allocated to each group will be. In this case, the size of each group (i.e., the number of allocated funnel information bits in each group) does not have to be the same, and the convergence speed of the BP decoding can be optimized to be maximized.

As shown in FIG. 12, when the polar BP decoding is performed using multiple CRCs, the decoding convergence speed can be improved and the decoding complexity can be reduced. There are two reasons for the reduced complexity. First, because convergence speed is faster, it is the first reason that it requires less calculation amount. The second reason is that the amount of computation can be reduced since there is no need to update the LLR value for information bits that have passed the CRC check.

CRC-based error correction

Conventional method

In general, the main reason for using CRC is to detect errors. The results of previous studies are as follows. In the following discussion, it is assumed that the CRC bits are added to the information bits and the CRC bits are directly transmitted to the channel (in the actual system, the CRC added information bits are polar encoded, Lt; / RTI &gt; channel).

The vector b = [b ₁ , b ₂ , ..., b _k ] ^T represents the information bits (b _i ∈ {0, 1}) and the vector r = [r ₁ , r ₂ , ... , r _m ] ^T denotes the CRC bits, and the vector c = [bT, rT] T = [C _l , C ₂ , ... , C _n ] represents the CRC code word (n = m + k) transmitted on the channel. Here, the CRC bit is designed so that the syndrome s ₀ becomes zero as shown in Equation (13).

Here, H is a parity check matrix corresponding to the CRC. The codeword c thus configured is modulated (actually after channel encoding such as polar encoding is performed) and transmitted on a wired or wireless channel. At the marine stage, the demodulation factor is performed (and channel decoding is performed) to obtain the received signal vector y. In the fishery, the LLR for the i-th coded bit is calculated using Equation (14) using y.

The final hard decision for each c _i from the LLR value is performed as: &lt; EMI ID = 15.0 &gt;

를 위에서 하드 결정된 모든 비트들을 포함하는 길이 n의 벡터라고 하면，하드 결정된 벡터

와 원래 전송된 코드워드 벡터 c와는 다음과 같은 관계가 성립한다.Here, vector

Is a vector of length n that includes all the bits hard-decided above,

And the originally transmitted codeword vector c satisfy the following relationship.

Here, e represents an error vector. At this time, the syndrome s ₀ can be easily given as shown in Equation (17).

를 알고 있으므로, 만약 수신단이 e를 알 수 있다면 수학식 16을 사용하여 원래의 코드워드 c를 추정할 수 있다. 즉 CRC를 사용하여 에러 정정을 수행할 수 있는 것이다. 그러므로 c를 추정하기 위해서는 e를 추정하는 것이 필요하다. At the receiving end

The receiver can estimate the original codeword c using Equation (16) if the receiver knows e. That is, error correction can be performed using CRC. Therefore, it is necessary to estimate e to estimate c.

It is possible to estimate the error pattern e having the maximum probability while satisfying the syndrome condition as in the expression (18).

Through some mathematical process, the above problem can be simplified as in Equation 19 (P1).

이며

은 가능한 2ⁿ 개의 모든 에러 패턴 중에서 주어진 신드롬을 만족하는 에러 패턴들의 집합으로, 수학식 20과 같다.only,

And

Is a set of error patterns satisfying a given syndrome among all 2 ⁿ possible error patterns.

는 길이 n인 2ⁿ개의 모든 가능한 에러 패턴의 집합을 나타낸다.here

Represents a set of all 2 ⁿ possible error patterns of length n.

In Equation (19), it is checked whether or not the syndrome condition is satisfied for all possible T error patterns, and an error pattern having the maximum probability among the T error patterns should be selected. In this case, there is a problem that the complexity is as high as 2 ⁿ .

, 이 비트들에 발생할 수 있는 길이 N_ur의 에러 패턴

를 추정하는 것이다. 이를 위해서, 먼저 길이 N_ur를 가지는

를 선택하는데, 이 벡터는 길이 n을 가지는 벡터

에서 가장 작은 값을 가지는 N_ur개의 원소를 선택하여 구축하는 벡터이다. 이 경우, 에러 추정 문제는 수학식 22 (P2)와 같이 주어진다. To solve the above complexity problem, we propose a simpler method. Instead of estimating the error pattern having the length n, only the N _ur bits having a relatively high probability of error occurrence due to the smallest absolute value of the LLR among the n bits are selected

, A length N _ur error pattern that can occur in these bits

. To this end, the first length having a N _ur

, Which is a vector having length n

Is a vector for selecting and constructing N _ur elements having the smallest value. In this case, the error estimation problem is given by Equation 22 (P2).

는 길이 N_ur의 에러 패턴을 나타내며,

은 수학식 13과 같이 정의 된다. only,

_Represents an error pattern of length N _ur ,

Is defined as &quot; (13) &quot;

는 가장 LLR의 절대값이 작아서 선택된 N_ur개의 정보 비트에서 발생할 수 있는 모든 N_ur개의 에러 패턴들의 집합을 나타낸다. In the above equation,

_Represents the set of all N _ur error patterns that can occur in the selected N _ur information bits since the absolute value of the LLR is small.

In Expression (22), the complexity is 2 ^Nr, which is lower than Expression (19). However, there is still a drawback that the complexity increases exponentially.

에서 가장 작은

값을 가지는

개의 원소들의 집합

를 매우 낮은 복잡도를 가지고서 선택해 낼 수 있다. 집합

는 수학식 25와 같이 정의된다. To overcome this problem, we propose a method to reduce the complexity. In this method, a set having 2 ^Nur elements

The smallest in

Value

A set of elements

Can be selected with very low complexity. set

Is defined as &quot; (25) &quot;

이며,

의 원소들에 대해서는 수학식 26의의 관계가 성립한다. here

Lt;

The relationship of the expression (26) holds.

For the given set as shown in Equation 26, error pattern estimation is performed as shown in Equation 27 (P3).

In the equation (27), the complexity is not given as an exponential function but given as an N _max value. In general, if N _max is selected as N _max = 3 N _ur , the error pattern estimation performance using Equation (27) It comes close to pattern performance estimation. In this case, the complexity when using equation (22) is given as 2 ^Nur , while the complexity when using equation (27) can be lowered to N _max = 3 N _ur .

In both error correction schemes using error patterns estimated using Equations 19, 22 and 27 given above, the performance of error correction is improved so that the length m of the CRC can be increased.

Suggested method

Hereinafter, the CRC-based error correction scheme described above is combined with HARQ. That is, when the packet is retransmitted due to the initial transmission failure, the receiver combines all the CRCs belonging to the plurality of received packets to perform CRC-based error correction, thereby maximizing the error correction performance.

Retransmission of packets (hard connection)

A packet of length n first including a CRC of length m is transmitted. The number of information bits is n-m = k. At this time, the syndrome calculated at the receiving end is expressed by Equation (28).

If s ₁ ≠ 0, it means that there is an error in the first packet. At this time, error correction is attempted through the existing CRC-based error correction methods (P1), (P2), and (P3). If the error correction fails, the same k pieces of information bits are retransmitted. The syndrome calculated at the receiving end is expressed by Equation (29).

In 2.2.1, it is assumed that the CRC used in the first packet and the CRC used in the second packet are the same.

을 첫번째 전송으로부터 수신단에서 하드 결정을 내린 코드워드,

를 두번째 전송으로부터 수신단에서 하드 결정을 내린 코드워드라고 한다. 이 경우, 실제 전송된 코드워드를 최적으로 추정하는 문제는 수학식 30과 같이 주어진다. Let vector c be the actually transmitted codeword,

Lt; RTI ID = 0.0 &gt; a &lt; / RTI &gt; hard word at the receiving end from the first transmission,

Is called a codeword with hard decision at the receiving end from the second transmission. In this case, the problem of optimally estimating the actually transmitted codeword is given by Equation (30).

와

을 고려하면, 수학식 31과 같은 관계가 성립함을 알 수 있다. Relationship between vectors

Wow

, It can be understood that a relationship as shown in Equation 31 is established.

이다.here,

to be.

을 최적으로 추정하는 방식은 수학식 32와 같다.After a little bit of mathematical process,

Is expressed by Equation (32).

이며,

,

,

으로 정의된다.here

Lt;

,

,

.

Comparing equations (32) and (19), it can be seen that both equations are given in the same form. In Equation (32), the length of the CRC is 2 m. Therefore, it can be seen that the use of Equation (32) improves the error correction performance as compared with Equation (19).

As in the prior art, the complexity of the present invention is also high as 2 ⁿ in the case of using Equation (32). In the case of applying the above-described method, the following complexity lowering mathematical expression is derived.

이다.here,

to be.

이다.here,

to be.

If the packet is transmitted K times, the above equation (32) can be expressed as Equation (35). At this time, it is assumed that all packets use the same CRC scheme.

,

,

,

이다.here,

,

,

,

to be.

If the packet is transmitted K times, the equation 33 is shown in Equation 36 ^{(P2 (2)),} Equation 34 is Equation 37 ^{(P3 (2)).}

이다.here,

to be.

이다.here,

to be.

Partial retransmission of packets (hard-coupled)

The first transmission transmits a packet of length n ₁ . The first packet contains k ₁ information bits and m ₁ CRC bits. In the second transmission, it is assumed that only the k ₂ information bits with the lowest reliability among the k ₁ information bits included in the first packet are selected and retransmitted. The second packet contains k ₂ information bits and m ₂ CRC bits (n ₂ ≤ n ₁ , k ₂ ≤ k ₁ ). In the case of polar coding, the reliability of the polar coding bit channel in which each bit is transmitted is different, and the different reliability values can be determined stochastically at the transmitting end. Also, the CRC used in the first packet and the CRC used in the second packet may be different.

At the receiving end, two syndromes can be obtained from two received packets as shown in equation (38).

Of the n ₁ columns of H ₁ , only k ₂ columns corresponding to the positions of k ₂ information bits transmitted in the second packet are selected to obtain a matrix H ^{`</sup> <sub>1</sub> having a size of m <sub>1</sub> xk <sub>2</sub> . Also, only k <sub>2</sub> columns corresponding to positions of k <sub>2</sub> information bits transmitted in the second packet among the n <sub>2</sub> columns of H <sub>2</sub> are selected to obtain a matrix H <sup>`} ₂ having a size of m ₂ xk ₂ . In this case, the syndromes obtained by the CRC parity check matrices satisfy the relation of Equation (39).

Here, e ' ₁ and e' ₂ represent the error patterns occurring in the first packet and the k ₂ information bits (i.e., the information bits transmitted by the second packet) in the second packet. In addition, the syndromes s ' ₁ and s' ₂ can be defined as shown in Equation (40).

However, where H ₁ '' is H in H 'is a matrix other than the H _2' 'is at the H ₂ H _2' is a matrix other than the. The vector e ₁ &quot; having the size k ₁ -k ₂ + m ₁ is an error in the first packet that combines errors occurring in k ₁ -k ₂ most reliable information bits and errors occurring in m ₁ CRC bits Pattern. The size m ₂ vector e ₂ '' is an error pattern that occurs at m ₂ CRC bits in the second packet. The two error pattern, e ₁ '' and e ₂ '' is in, but not known to the receiving end, the purpose of this invention in the bits under the assumption that the probability of an error is small _{e 1 '' = 0, e} 2 ''= 0 is assumed. Since the receiving end knows s ₁ , s ₂ , H ₁ '', and H ₂ '', s ₁ 'and s ₂ ' can be calculated.

(Mathematical Expression 41 (P1 ⁽³⁾ ) used in the error correction method can be derived through the above-described mathematical process.

,

,

,

이다.here,

,

,

,

to be.

When the above-described method is applied, a mathematical expression with reduced complexity is derived as shown in Equation 42 (P2 ⁽³⁾ ).

이다.here,

to be.

이다. here,

to be.

At this time, it is assumed that K packets are transmitted. The length of a packet in each transmission is n _k (k = 1, 2, ..., K) and the following relation holds.

n ₁ ? n ₂ ? ≥ n _K

In addition, the number of information bits included in each packet has the following relationship.

k ₁ ? k ₂ ? ≥ k _K

It is assumed that all packets can have different CRCs of length m _k (k = 1,2, ..., K).

In this case, the transmission method (P1 ⁽⁴⁾ ) can use equation (44).

,

,

,

이다.here,

,

,

,

to be.

As in the foregoing, Equation 46 can be used for the complexity lowering (P2 ⁽⁴⁾ ) and the lower complexity (P3 ⁽⁴⁾ ).

이다.here,

to be.

이다.here,

to be.

Packet full retransmission: soft combining

When the entire packet is retransmitted, the receiving end combines the soft value (LLR value) of the received packet and proposes a method of estimating the error pattern. The received signal vector for the _kth received packet at the fishery end is denoted by y _k . At the receiving end, {y _k : k = 1, 2, ... , K}, the i-th LLR for the combined packets is calculated as shown in equation (46).

이다.here,

to be.

The final hard decision for each c _i from the LLR values is performed as shown in equation (48).

는 하드 결정된 모든 모든 비트들을 포함하는 길이 n의 벡터이다.

는 모든 패킷의 LLR이 결합된 값을 나타내는 벡터이다.vector

Is a vector of length n containing all the bits that are hard-determined.

Is a vector representing the combined value of the LLRs of all packets.

를 사용하여 송신된 패킷을 추정한다. 이와 같이 추정된 패킷에 대하여 CRC체크를 수행하여 CRC 체크가 성공한 경우 에러를 정확히 추정했다고 결론을 내린다.The error correction can be performed by superposing these values on the existing methods (Pl), (P2), and (P3). The method given in this way is represented by (Pl ⁽⁵⁾ ), (P2 ⁽⁵⁾ ), and (P3 ⁽⁵⁾ ). In this case, by estimating the error pattern e for the combined packet,

To estimate the transmitted packet. It is concluded that the CRC check is performed on the packet so estimated, and the error is accurately estimated when the CRC check is successful.

Packet Part Retransmission: Soft Combination

Even when only a part of the packet is retransmitted, soft combining can be performed and the error estimation can be performed in the manner described in Section 2.2.3. One difference is that when only a portion of the packet is transmitted, the LLR value is added only for the retransmitted information bits. Otherwise, it is the same as that proposed in Section 2.2.3. ⁽⁶ ), ⁽⁶⁾ , and (P3 ⁽⁶⁾ ) show how to perform error estimation using soft combining in the case of partial retransmission of a packet.

2.2.5 Combining multi-CRC applied error correction and polar coding-based HARQ using multiple CRC

FIG. 13 shows a method of performing error correction by applying multiple CRCs and a method of combining HARQ-based data retransmission based on polar coding based on multiple CRCs. 13, Z represents a frozen bit used for polar coding. The first packet consists of two mini-packets, and assumes that CRC ₁ failed when the first packet was transmitted. At this time, the receiver first performs error correction based on (P1), (P2) or (P3) using CRC1. If the error correction is unsuccessful, not only the entire first mini packet which has been confirmed to have an error is retransmitted but only CRC ₃ is added to some bit group a ₁ transmitted in a low-reliability polar bit channel among the mini- . This is based on the binary search method described above.

If the CRC ₃ fails again on the basis of the two CRC using (that is, CRC ₁ to ^{_{CRC 3) (P1 (3)}} ), (P2 (3)), (P3 (3)) performs error correction . If such error correction fails, a part of the bit group a11 transmitted with a low-reliability polar bit channel is transmitted again at a1.

If CRC ₃ succeeds, it is confirmed that there is no error in a1 received in the second packet, and this is substituted into a1 of the first packet. Now CRC1 is checked again, and if it fails, there is an error in a2. Now, CRC ₁ is used to perform error correction based on (P1), (P2), and (P3). If such error correction fails, a part of the bit group a21 transmitted with a low-reliability polar bit channel is again transmitted at a2.

If both CRC ₃ and CRC ₁ are successful, the next transmission scheme will be different depending on the result of CRC ₂ . If CRC ₂ fails, repeat the above procedure for b1 and b2. If CRC2 succeeds, it sends a completely new packet.

3. Resolution of the undetected error probability (UEP) problem in the CRC.

Generally, the CRC is effective in detecting the presence of errors, but the error detection probability is not one. The probability that the CRC does not detect the existence of this error is actually called the undetected error probability (UEP), and the following approximation formula is often used in the literature.

Here, m is the number of bits used for CRC. At this time, as the length of m increases, the UEP exponentially decreases, and when m decreases, the UEP exponentially increases.

3.1 Error detection using dual layer CRC

To solve this UEP problem, it is possible to use CRC in two layers. As shown in FIG. 14A, CRC ₁ , CRC ₂ , and CRC ₃ are CRCs used in the first layer, which perform error detection of each mini-packet. CRC ₀ is the CRC used for the second layer. CRC ₀ acts to detect errors occurring in any of the three mini-packets. As a result, errors occurring in each mini-packet are detected double by CRC _i , i = 1,2,3 and CRC ₀ . Hereinafter, an efficient retransmission method and an error correction method using CRC are developed when the CRC is used as a double layer as shown in FIG. 14A.

(i=1,…, L)는 수신단에서 수신되어 하드 결정이 내려진 L개의 미니 패킷을 나타내며, 각 미니 패킷 끝에는 한개의 CRC가 포함되어 있다. 각 미니 패킷의 길이는 ni이다. 길이 n의 벡터

는 모든 미니 패킷을 포함한 패킷을 나타내며, 패킷의 끝에는 추가로 한개의 CRC가 더 연결되어 있다.vector

(i = 1, ..., L) denotes L mini packets received at the receiving end and subjected to hard decision, and one CRC is included at the end of each mini-packet. The length of each mini-packet is ni. Vector of length n

Indicates a packet including all mini-packets, and an additional CRC is further connected to the end of the packet.

,

이고, 수신단은 하기 수학식 50과 같이 미니 패킷들에 대한 신드롭들과 패킷에 대한 신드롬을 얻게 된다.here,

,

, And the receiving end obtains a new drop for the mini packets and a syndrome for the packet as shown in Equation (50).

The receiving end obtains syndromes for the mini packets and a syndrome for the packets as follows.

Here, H _i (i = 1, 2, ..., L) is the parity check matrix of the CRC used in each mini-packet of the CRC used in each mini-packet, and $ \ bfH_0 $ is the parity check matrix to be.

When CRC is used as the double layer, the UEP is given by the following equation (51).

Where P _UEP (0) is the UEP for CRC0, and P _UEP (i) is the UEP for CRCi. Ultimately, using a CRC as a double layer can reduce the UEP much more than using a single layer of CRC.

3.1.1 Retransmission method with dual-layer CRC

When data is transmitted using a dual layer CRC, the following three cases can be considered at the receiving end.

새로운 패킷을 전송한다.

And transmits a new packet.

에러가 발생한 미니 패킷(들)을 알 수 있으므로, 그 미니 패킷(들)이 재 전송 된다.

Since the minipack (s) in which the error occurred can be known, the minipack (s) are retransmitted.

이 경우에는, 패킷 내에서 에러가 발생한 것은 알지만, 정확히 어느 미니 패킷에서 에러가 발생했는지는 알지 못한다. 즉, 미니 패킷에서 사용된 CRC들이 에러 탐지에 실패한 경우이다. 이 경우에 대한, 재 전송 방식을 아래에서 논의한다.

In this case, it is known that an error has occurred in the packet, but it is not known exactly which error occurred in which mini-packet. That is, the CRCs used in the mini-packet fail to detect errors. The retransmission scheme for this case is discussed below.

In one embodiment, the entire packet may be retransmitted. However, such a retransmission scheme is very inefficient, especially when the value of L is large. To solve this problem, the following method can be used.

First, the L mini-packets are listed so that the probability that an error occurs in each mini-packet is smaller than the probability of occurrence of an error. That is, they are listed in order from smallest to least likely to have not caused errors. The index of the mini-packet listed is i1, i2, ... , i _L.

은 수학식 53과 같다.The probability that an error will not occur in the lth mini packet

Is expressed by Equation (53).

At the transmitting end, one mini packet is transmitted i ₁ , i ₂ , ... , i _L. At the receiving end, CRC of two layers is performed whenever one mini packet is received. If the CRC check is successful, the transmitting end transmits a completely new packet. If the CRC check fails, the next mini-packet is retransmitted in order.

3.1.2. Error correction method using CRC when using dual layer CRC

The following three cases can be considered.

새로운 패킷을 전송한다.

And transmits a new packet.

Since the mini packet (s) in which an error occurs can be known, error correction is performed using CRC of two layers.

If the error correction fails, the mini-packet in which the error occurred is retransmitted.

It is not possible to know the mini packet (s) in which the error occurred. In this case, a method of performing error correction using the CRC is required, and this method is described below.

If the error correction fails, the mini-packet is retransmitted one by one based on the probability in the manner described in 3.1.1.

First, the mini-packet is sorted as shown in Equation (54).

Now, starting from the mini packet corresponding to i ₁ , i _l , l = 1, 2, ... , K, the error correction based on the CRC is performed using the following equation (55 (P2 ⁽⁷⁾ ).

,

,

이다.here

,

,

to be.

Where H _{0, il} is structured to select from a matrix having a size m ₀ xn _il, the column that indicates the bit position corresponding to the i-th mini packet H _{0 l.} In order to reduce the amount of calculation, it is possible to use the same method as in Equation (56) (P3 ⁽⁷⁾ ) below.

이다.here,

to be.

3.2 Partially overlapping CRC transmission scheme

In the above, a method of using CRC as a double layer is proposed. This approach is very effective in reducing UEP, but it has the disadvantage that CRC should be used too much. Therefore, in the following description, one CRC is used for each mini-packet, but a CRC is not used for the packet. Conventional in FIG. 14B shows an existing method. As shown in FIG. 14B, one CRC is used for each mini-packet, but no CRC is used for the entire packet. 14B shows a CRC scheme that partially overlaps a CRC proposed by the present invention. For this scheme, we propose new methods for error detection, retransmission, error correction, and polar decoding below.

3.2.1 Error detection

Compared with the conventional method, the proposed method shows better error detection performance. For example, in FIG. 14B, the first mini-packet a is detected by only one CRC in the conventional scheme, but is detected by three CRCs in the proposed scheme. Therefore, in the case of the proposed scheme, the UEP for the first mini-packet becomes very small. In the case of the second mini-packet b, only one CRC is detected in the conventional scheme, but in the proposed scheme it is detected by two CRCs. Therefore, the UEP for the second mini-packet is also reduced. In the case of the last and third mini-packets, the existing scheme and the proposed scheme have the same UEP.

3.2.2 Retransmission

When transmitting using the partially overlapped CRC proposed in FIG. 14B, the CRC check of CRC ₁ and CRC ₂ is passed at the receiving end, but the CRC check of CRC ₃ may fail. In this case, theoretically, it can be seen that an error occurred in one of the mini-packets a, b, and c. However, the probability that an error occurred is different. First, the probability that an error occurred in the first mini-packet is the smallest. This is because the probability that CRC ₁ can not detect the error of the first mini-packet, the probability that CRC ₂ can not detect the error of the first mini-packet, and the probability that CRC ₃ detects the error of the first mini-packet . On the other hand, the probability of an error in the second mini-packet is higher. For the same reason, the probability of an error in the third mini-packet is highest. In the retransmission method proposed in the present invention, the third mini packet having the highest probability of occurrence of an error is first retransmitted. Then check CRC ₃ again. If unsuccessful, retransmit the second mini-packet, and if CRC ₃ fails again, retransmit the first mini-packet.

The above proposed method can be applied even when K CRCs are overlapped. For example, if all of CRC ₁ to CRC _K-1 are successful, and CRC _K fails, the K-1th mini-packet starting from the K-th mini packet, ... , And retransmits the mini-packet in the order of the first mini-packet.

3.2.3 Error Correction

Instead of the proposed retransmission, error correction using CRC can be performed. When determining the order of error correction, it is considered that the probability that an error occurred in each mini-packet is different. That is, the error correction based on the CRC starts from the K-th mini packet, the K-1-th mini packet, ... , And the first mini-packet.

3.2.4 Polar decoding

The superimposed CRC proposed in the present invention can also be applied to polar decoding using BP decoding. This concept is shown in Fig. 14C. When performing polar decoding using BP decoding, check each CRC and if the CRC succeeds, set the LLR of the information bits corresponding to the CRC to either positive infinity or negative infinity. In this way, BP decoding can be performed more efficiently. At this time, the UEP of the CRC may be a problem. This UEP problem can be largely solved by using the partially overlapping CRC proposed in the present invention.

3.2.5 Modified partially collapsed CRC transmission scheme

Several modified partially overlapping CRC transmission schemes are conceivable. 14D shows one variant. This modified manner can be relatively bad for the UEP performance as compared to the proposed scheme given in FIG. 14D. However, since the number of information bits of each CRC is small, the amount of calculation required for the CRC check is small. As a result, when the left and right schemes of FIG. 14D are regarded as two extremes, it is possible to consider a number of modified schemes in the middle.

4. Non-systematic polar code and systematic polar code

There are two kinds of polar codes. One is a non-systematic polar code and the other is a systematic polar code. 15 shows an unstructured polar code and a systematic polar code. First, the encoding of the polar code will be described below.

가 생성 행렬에 곱해 지는 (즉, 입력되는) 비트들의 벡터라고 하자. 또한,

는 생성 행렬에 곱하여 얻어지는 (즉, 출력되는) 비트 들의 벡터라고 하자. 실제로 채널로 전송되는 코드워드 (codeword)는 x이다. 위에서

는 정보 비트가 할당되는 위치를 나타내는 인덱스들의 집합이다. The polar code is encoded by a generator matrix G.

Is the vector of bits that are multiplied (i.e., input) by the generator matrix. Also,

Is the vector of bits obtained (i.e., output) by multiplying the generator matrix. The codeword actually transmitted to the channel is x. Above

Is a set of indices indicating the locations where information bits are allocated.

에 할당되고 정보 비트가

에 할당된다. 흔히, 정보 비트의 마지막에는 CRC (cyclic redundancy check) 비트가 첨부된다. x=uG에 의해서 코드워드가 얻어지며 이것이 채널로 전송된다. In non-systematic polar code,

And the information bits are

Lt; / RTI &gt; Often, a cyclic redundancy check (CRC) bit is appended to the end of the information bits. A code word is obtained by x = uG, which is transmitted on the channel.

에 할당되고 정보 비트가

에 할당된다. 흔히, 정보 비트의 마지막에는 CRC (cyclic redundancy check) 비트가 첨부된다.

와

를 선행 문헌에 기재된 바와 같이 결정한 후, 코드워드

를 채널로 전송한다. In systematic polar code, frozen bits

And the information bits are

Lt; / RTI &gt; Often, a cyclic redundancy check (CRC) bit is appended to the end of the information bits.

Wow

Is determined as described in the prior art, then the codeword

To the channel.

의 디코딩된 값인

를 얻는다.The decoding of the unstructured polar code is performed as follows. There are various decoding schemes. The most common method is successive interference (SC) decoding. List SC decoding may be used to improve its performance, and some of the information bits may be used for CRC coding to further enhance its performance. There are BP (belief propagation) schemes for obtaining soft values by other decoding methods. The method proposed by the present invention can be decoded by any decoding method. The decoding scheme may be List SC decoding. In decoding non-systematic polar codes, using this decoding scheme

&Lt; / RTI &gt;

.

를 얻는다. 두번째 단계에서는 재 인코딩 (re-encoding)의 단계를 거쳐서

를 얻는다. 체계적 폴라 코딩의 정보 비트의 디코딩 값인

는 재 인코딩에 의해서 구축된 코드워드

에 포함되어 있다.In the case of decoding systematic polar codes, decoding takes two steps. In the first step, as in unstructured polar coding

. In the second stage, the re-encoding step

. The decoded value of the information bits of systematic polar coding

Lt; RTI ID = 0.0 &gt; re-encoding &lt; / RTI &

.

The systematic polar coding and non-systematic polar coding have the same frame error rate (FER). However, the bit error rate (BER) performance is better than systematic polar coding. That is, systematic polar coding at the same signal to noise ratio (SNR) shows a lower BER than non-systematic polar coding. The reason for this is due to the error attenuation effect obtained in re-encoding, which is the second step of decoding systematic polar coding.

A method of combining polar coding and hybrid automatic repeat request (HARQ) may be considered in order to correspond to an error that occurs when a data packet encoded by a polar code is transmitted to an actual channel. There are methods that combine non-systematic polar code and HARQ. In the HARQ scheme, a code rate of a first packet to be transmitted is set high. If the transmission of the first packet is not successful, the code rate of the overall viewpoint is lowered through retransmission. If the code rate is low, the error probability is also low. Therefore, as the retransmission is repeated, the code rate is lowered and the probability that the packet is successfully decoded is also increased. The higher code rate of the first packet in the HARQ scheme assumes that the channel is sufficiently good at the time of transmission of the first packet. When setting the code rate of the first packet, if the code rate is set too high, the probability of the first packet failing to decode will be too high and retransmission should occur with high probability. In this case, a transmission delay occurs. If the code rate of the first packet is set too low, there is a disadvantage in that, when the actual channel is very good, all of the large channel capacity provided by the good channel can not be used.

There is a multi-layer transmission (MLC) scheme as an approach to compensate for the drawbacks of such HARQ. Unlike HARQ, multi-layer transmission has several layers in one packet and transmits information bits to each layer [Reference 2], [Reference 3]. If the channel is bad, only the information bits of the first layer can be decoded. However, if the channel is good, the next layer of information bits can be decoded. If the channel is very good, it can decode information bits of all layers. Multilayer transmission is often referred to in the literature as superposition coding (SPC). However, multi-layer transmission is different from commonly known superposition coding. In multi-layer transmission, it is used when there is one transmitter and one receiver. When there are two layers as an example of the multi-layer transmission, the transmission signal x and the reception signal y are given by Equation (57).

Where P ₁ + P ₂ = P _T. In the above equation, s _i represents the signal transmitted to the i th layer, Pi represents the power allocated to the i th layer, and P _T represents the total power. The channel capacity of the first layer and the second layer is given by Equation (58).

As described above, when data is transmitted using multiple layers, the throughput or the average data rate in the fading environment can be improved. Although the present invention can be applied to a multi-layer transmission scheme having any number of layers of the scheme proposed in the present invention, for the sake of discussion, only two layers are assumed below.

In the past, there was no system that combined systematic polar coding and HARQ. Systematic polar coding shows better BER performance than non-systematic polar coding, so this research is indispensable.

In addition, a scheme of combining systematic polar coding and HARQ with multi-layer transmission has not been developed. HARQ and multi - layer transmission have a complementary relationship with each other, and research using a systematic polar code having good performance is indispensable.

4.1 HARQ using one CRC per packet

4.1.1 HARQ 1s scheme

는 코딩되어 전송되는 정보 비트의 집합을 나타낸다. 이 정보 비트의 집합 중에서 가장 신뢰도가 낮은 정보 비트들로만 구성된 집합은

로 나타낸다. 이와 같이

이

로 부터 생성되는 경우

로 나타낸다.The first scheme proposed in the present invention is to use a systematic polar code for all packets. This scheme is called HARQ 1s scheme in the present invention. Where s represents a single CRC. 16A shows the HARQ 1s scheme. As can be seen in Figure 16a, the first packet is coded and transmitted by a systematic polar code. If the CRC check result at the receiving end fails, this is known to the transmitting end, and the transmitting end transmits the second packet. Of the first packet

Represents a set of information bits that are coded and transmitted. A set consisting of only the least reliable information bits among the set of information bits

Respectively. like this

this

&Lt; / RTI &gt;

Respectively.

Both the first packet and the CRC attached to the second packet are coded by the systematic polar code.

를 전송한다. 두번째 전송도 실패한 경우에는, 송신단에서 첫번? 전송된 비트들의 신뢰도와 두번째 선택되어 전송된 비트들의 신뢰도를 합하여, 전송된 비트 들의 신뢰도를 새로 업데이트 한다. 이러한 과정에 의하여 정보 비트들의 새로 업데이트된 신뢰도를

로 나타낸다. 이제

에서 가장 낮은 신뢰도를 가지는 정보 비트들을 선택하여, 세번째 전송에서 전송하게 된다. 즉, 세번째 전송에서 전송되는 비트들의 집합

는

로 나타낼 수 있다.16B shows a case where retransmission occurs twice (all three transmissions including the initial transmission). If the first transmission fails

. If the second transmission also fails, The reliability of the transmitted bits and the reliability of the second selected transmitted bits are combined to update the reliability of the transmitted bits. By this process, the newly updated reliability of the information bits

Respectively. now

And selects the information bits having the lowest reliability in the third transmission. That is, the set of bits transmitted in the third transmission

The

.

In addition, the CRC bits of all packets are coded and transmitted by systematic polar coding.

이고 두번째 패킷의 첫 단계에서 디코딩되어야 하는 비트는

이다. 이 때,

에 있는 비트들과

에 있는 비트들은 일반적으로 다른 값들을 가지므로, 결합 리스트 SC 디코딩을 사용하여 디코딩을 수행할 수 없다. HARQ 1s 방식의 두번째 제한점은, 정보 비트의 집합

,

,

,

을 계산하기 어렵다는 것이다. 이 값들은 송신단에서 계산되어야 하는데, 현재까지 체계적 폴라 코드에 대하여 그러한 정보 비트 집합 내의 각 정보 비트의 신뢰도를 (시뮬레이션이 아닌) 해석적인 방법으로 계산하는 방법은 알려져 있지 않다.16A and 16B, there are two limitations in the case of the HARQ 1s scheme. The first limitation is decoding. It is possible to perform decoding using an incremental freezing scheme in which the retransmitted packets are individually decoded and substituted into the previous packet. However, combining list SC decoding is not possible. The reason for this is as follows. The first step in decoding the systematic polar code is to perform the same decoding as the unstructured polar code. However, in the HARQ 1s scheme, the bits to be decoded in the first stage of the first packet are

And the bits to be decoded in the first stage of the second packet are

to be. At this time,

With the bits in

Can not be decoded using combined list SC decoding since the bits in &lt; RTI ID = 0.0 &gt; A second limitation of the HARQ 1s scheme is that a set of information bits

,

,

,

It is difficult to calculate. These values have to be calculated at the transmitting end. To date, it is not known how to calculate the reliability of each information bit in such a set of information bits for a systematic polar code in an analytical way (not a simulation).

4.1.2 HARQ 2s scheme

이 결정되면, 두번째 패킷에서는

중에서 가장 신뢰도가 낮은 비트들만을 선택하여

을 구축하며,

에 속한 비트들을 비 체계적 폴라 코딩을 사용하여 인코딩하여 전송한다. 즉,

의 관계가 성립한다.16C shows the HARQ 2s scheme as a second proposed scheme. In this way, only the first packet is encoded by the systematic polar code. Through this encoding process

Is determined, in the second packet

Only the least reliable bits are selected

In addition,

And then transmits the encoded bits by using non-systematic polar coding. In other words,

.

The CRC of the first packet is encoded by systematic polar coding, while the CRC of the second packet is encoded by non-systematic polar coding.

과 두번째 패킷의

를 결합하여 신뢰도를 업데이트 한 뒤, 이 중에서 신뢰도가 가장 낮은 비트 들만을 선택하여

를

와 같이 구축한다.16D shows an example in which two retransmissions (three transmissions combining the first transmissions) occur in the HARQ 2s scheme. Of the first packet

And the second packet

To update the reliability, and then only the least reliable bits are selected

To

.

내에 있는 정보 비트들을 비 체계적 폴라 코딩을 사용하여 인코딩 한 후 전송한다.In the third packet

Encoded using non-systematic polar coding and then transmits the information bits.

The CRC of the first packet is encoded by systematic polar coding, but the CRCs used in the retransmission are encoded and then transmitted by non-systematic polar coding.

를 디코드 한 후, 한 번의 재 인코딩을 통하여

를 구하게 되는데, 이와 같은 재 인코딩 단계에서 에러 감쇄의 효과가 나타나게 된다.The advantage of the HARQ 2s scheme is that it is possible to perform decoding in the combining list SC scheme as well as in the progressive freezing scheme at the receiving end. FIG. 16E shows that combined list SC decoding is possible. In this manner, the combined list decoding is performed

Decoded, and then re-encoded once

The effect of error reduction is exhibited in the re-encoding step.

,

,

,

에 속한 비트 들의 신뢰도를 기존의 비 체계적 폴라 코딩에서와 마찬가지로 density evolution (DE) 이나 가우시안 근사 (Gaussian approximation) 을 통하여 쉽게 구할 수 있다는 것이다. 예를 들어,

의 신뢰도 값인 log likelihood ratio (LLR)값들은 다음 수학식 59와 같이 간단히 계산되어 질 수 있다.A second advantage of the HARQ 2s scheme is that,

,

,

,

(DE) or Gaussian approximation (Gaussian approximation) as in the conventional non-systematic polar coding. E.g,

The log likelihood ratio (LLR) values, which are the confidence values of the log likelihood ratio (LLR), can be simply calculated as follows.

4.2 HARQ using multiple CRCs per packet

When constructing the HARQ of polar coding, it is conceivable to use a plurality of multiple CRCs. In this scheme, the information bits of the first packet are divided into several groups and each group is appended with one CRC. As a result, the first packet is transmitted using a plurality of CRCs, and the receiver notifies the transmitter of the number of the CRC that failed the CRC check. The transmitting end transmits the information bits that the CRC failed to CRC check. At this time, the information bits to be retransmitted are divided into several small groups again, and CRC is added to each group. If the receiver fails the CRC check after performing the CRC check, it notifies the transmitter of the number of the failed CRC. At the transmitting end, the information bits that the failed CRC was divided into are divided into several smaller groups, and the CRC is added, and the polar coded is performed to transmit. When HARQ is performed in this manner, as the retransmission is further performed, the size of the information bit group handled by each CRC becomes smaller. Therefore, it is possible to estimate the position of the information bit issued by the error through the minimum transmission have. This method is called binary search (or bi-section method). In the following, a systematic polar coding-based HARQ scheme using multiple CRCs is proposed.

4.2.1 HARQ 1m scheme

The first scheme is called the HARQ 1m scheme, and its concept is shown in FIG. 17A. The HARQ 1m scheme is a scheme in which multiple CRCs are used in the HARQ 1s scheme. Where m indicates that multi-CRC is used. This scheme has the advantage that all information bits and all CRCs are encoded and transmitted by systematic polar coding. However, the disadvantage is that while decoding is possible by progressive freezing method, decoding of binding list SC is not possible. Another disadvantage is that the use of multiple multiple CRCs can reduce the decoding complexity as described in the existing literature, but the HARQ 1m scheme can reduce the decoding complexity even though multiple multiple CRCs are used It is not.

4.2.2 HARQ 2m scheme

에 사용되었지만, HARQ 2m 방식에서는 여러 CRC가 최초의 패킷의 정보 비트가 아닌 부호화된 비트들 {c_k:k=1,2, …, 6}에 사용된다는 것이다. HARQ 2m 방식에서도 첫번째 패킷의 정보 비트

에 대하며 CRC를 사용할 수 있다 (도 17b의 CRC₄). 하지만. 이와 같이

에 대하며 CRC를 사용하는 것이 반드시 필요한 것은 아니다. 그럼에도,

에 대하여 CRC를 사용하면, 첫번째 패킷에 대해서는 이 중으로 CRC가 사용되므로, CRC가 에러 탐지에 실패하는 확률을 줄일 수 있다.The HARQ 2m scheme is extended by using multiple CRC in the HARQ 2s scheme. FIG. 17B shows this method. However, HARQ 2m is not a very simple extension of HARQ 2s. In the HARQ 2s scheme, in the first transmission packet, the CRC is an information bit

, But in the HARQ 2m scheme, multiple CRCs are used for coded bits {c _k : k = 1,2, ..., , 6}. Even in the HARQ 2m scheme,

And CRC can be used (CRC _{4 in} FIG. 17B). But. like this

It is not necessary to use CRC. Nevertheless,

The CRC is used for the first packet, so that the probability that the CRC fails to detect an error can be reduced.

The greatest advantage of the HARQ 2m scheme is that not only the progressive freeing scheme but also the combined list SC decoding is possible at the time of decoding. In addition, the complexity / calculation amount reduction effect can be obtained in the decoding that can be obtained by using multiple CRCs.

4.2.3. HARQ 3m method

FIG. 17C shows a third possible method and is called HARQ 3m scheme. In the first packet, the information bits {a _k : k = 1, 2, ... , 6} are divided into several groups, and one CRC is added to each group. The information bits and all CRC bits are encoded and transmitted by systematic polar coding. The receiver performs a check on all CRCs and transmits the failed CRC number to the transmitter. In the second transmission, the information bits not successfully decoded in the first transmission and the parity check portion in the codeword transmitted in the first transmission are transmitted together. In the same way, a third transport packet is also constructed. 17D shows transmission of the HARQ 3m scheme, for example, of a polar code having an initial packet length of 8.

The HARQ 3m scheme has several advantages. First, the first advantage is that, at the transmitting end, encoding is not required at all retransmission steps except the first packet, so the amount of computation at the transmitting end is not high. An important advantage is that, in decoding at the receiving end, not only the progressive free decoding method but also the combined list SC decoding is possible. Lastly, in the HARQ 3m scheme, when performing combined list SC decoding using multiple CRCs, decoding complexity can be reduced by multiple CRCs.

17E shows decoding. In the first stage of decoding, the combined list SC decoding is performed by combining the first, second and third packets. At this time, the information bits {a ₁ , a ₂ , a ₅ , a ₆ } that were successfully decoded in the _first packet are combined into the second packet so that the LLR values in the second packet are generated. Likewise, the information bits {a ₁ , a ₂ , a ₃ , a ₅ , a ₆ } that were successfully decoded in the first and second packets are combined into a third packet such that the LLR values in the third packet are generated. Now, the LLR calculated in the first packet, the LLR value generated from the second packet with the help of the first packet, and the LLR value of the third packet generated with the help of the first packet and the second packet are added together to perform SC decoding do. Finally, re-encoding is performed.

Another method of decoding the HARQ 3m scheme is to merely combine the received signals at the receiving end through a maximal ratio combining (MRC) scheme, and then perform list SC decoding.

5. Multi-layer transmission based HARQ using systematic polar code

5.1 FER and BER of information bits in polar coding

If the transmitted packet is not completely decoded, the CRC check will fail. In this case, a frame error occurs. The probability of occurrence of a frame error is FER (frame error rate). Now we look at the relationship between FER and BER. If the length N of the code is finite, the relationship between FER and BER can be represented by a union bound as shown in Equation 60. [

에 대하여 측정된 확률을 나타내고, 체계적 폴라 코딩의 경우에는

에 대하여 측정된 확률을 나타낸다. 여기서 중요한 점은 프레임 에러가 발생한 상황에서도 패킷 내에 실제로 존재하는 비트 에러의 갯수는 일반적으로 크지 않다. 예를 들어, 프레임 에러는 한개의 비트 에러에 의해서도 발생된다.Union bounds are more accurate when the channel's signal-to-noise ratio (SNR) is high. As can be seen from this formula, the BER value is generally much smaller than the FER value. In the case of actual polar code, the value of BER is much smaller than FER. In this case, BER and FER are used for non-systematic polar coding

, And in the case of systematic polar coding,

Lt; / RTI &gt; The important point here is that the number of bit errors actually present in the packet is generally not large even in the presence of a frame error. For example, a frame error is also caused by a single bit error.

의 BER뿐 아니라 페러티 체크 비트

도 작아진다. 결국, 정보 비트와 페러치 첵 비트로 이루어 진 코드 워드

의 전체 BER이 작은 것이다. 재 구축된 코드워드

의 BER이 높지 않으므로, 이 코드워드

내에 존재하는 비트 에러의 갯수는 많지 않다. 결과적으로, 체계적 폴라 코드를 사용하여 다중 계층 전송을 구현하면, 특정한 계층에 프레임 에러가 발생한 상황에서도, 이 계층이 다른 계층에 미치는 간섭은 일반적으로 크기 않다.Inter-layer interference in a multi-layer transmission depends on how many bit errors are present in the codeword (reconstructed after decoding). Therefore, the bit error we are interested in is the BER of the codeword. In a systematic polar code,

The BER of the parity check bit

. As a result, the code word consisting of the information bits and the parity check bits

Lt; / RTI &gt; Reconstructed codeword

The BER of the codeword is not high,

The number of bit errors present in the bit stream is not large. As a result, implementing multi-layer transmission using systematic polar codes will not generally cause interference to other layers of this layer, even in the presence of frame errors in certain layers.

5.2 iterative decoding between layers in a multi-layer transmission

In the previous section, we have found that there is generally a small number of bit errors in the reconstructed codeword, even if frame errors occur, when using systematic polar codes. With this effect, if a systematic polar code is used for multi-layer transmission, decoding can be attempted in the second layer even if a frame error occurs in the first layer. In this section, we further refine this concept and propose an iterative decoding scheme between layers in multi-layer transmission.

FIG. 19A shows a scheme of using interleaved iterative decoding in the double layer transmission. FIG. 19B shows a concept of a scheme of using interleaved iterative decoding in the double layer transmission. By performing inter-layer decoding repeatedly, even when a frame error occurs in one layer, decoding is performed in the remaining layers, and interference between layers is reduced as decoding is performed. This is because the signal-to-interference and noise ratios signal to interference plus noise ratio (SINR), resulting in a higher probability of successful decoding.

5.3 Transmission methods

5.3.1 HARQ 1s-MLT scheme

20A is a method of combining HARQ 1s (FIG. 16A and FIG. 16B) proposed in Section 4.1.1 and multi-layer transmission (MLT), and is referred to as HARQ 1s-MLT. The decoding is performed as follows. For each layer, decoding is performed in a progressive freezing method. In addition, over the two layers, the iterative decoding between the layers proposed in the previous section is performed.

나

를 결정하기 위해서는, 체계적 폴라 코딩에서 각 정보 비트의 신뢰도를 결정해야 하는데, 이러한 결정을 간단히 하는 방법이 알려져 있지 않다.The HARQ 1s-MLT scheme inherently has limitations of the HARQ 1s scheme. For example, combining list SC decoding can not be used for each layer. Also,

I

It is necessary to determine the reliability of each information bit in systematic polar coding. However, a method of simplifying such a determination is not known.

5.3.2 HARQ 2s-MLT scheme

FIG. 20B shows a HARQ 2s-MLT scheme in which a HARQ 2s scheme (FIGS. 16C and 16D) proposed in Section 4.1.2 is combined with a multi-layer transmission (MLT) scheme. The decoding is performed as follows. For each layer, it performs joint list SC decoding. In addition, over the two layers, it performs iterative decoding between the layers proposed in the immediately preceding section.

Combination list SC decoding + Iterative decoding between layers

In the first stage of decoding, a binding list SC decoding is performed using several packets of the first hierarchy to estimate the first packet of the first hierarchy.

을 재 인코딩하여 코드워드

를 생성한다.In the second phase,

Lt; RTI ID = 0.0 &gt; codeword &lt;

.

In the third step, we perform iterative decoding between layers.

나

를 결정하기 위하여 필요한 비트의 신뢰도를 쉽게 계산할 수 있다는 것이다.Another advantage of the HARQ 2s-MLT scheme is that

I

The reliability of the bits required to determine the bit error rate can be easily calculated.

5.3.2 HARQ 1m-MLT method

20C is a method of combining the HARQ 1m scheme (FIG. 17A) proposed in Section 4.2.1 and a multi-layer transmission (MLT) scheme, and is referred to as HARQ 1m-MLT scheme. This scheme has advantages and disadvantages of the HARQ 1m scheme except that it increases the transmission rate by multi-layer transmission. Decoding is possible for progressive free decoding, but binding list SC decoding is not possible. In addition, even though multiple CRCs are used, a reduction in decoding complexity can not be obtained. We perform the inter-layer iterative decoding between layers. Because such decoding is used, the bi-section method (or binary search method) is used not only in the first layer but also in the second layer, and as the retransmission is repeated, the location where the error occurs quickly can be found.

5.3.4 HARQ 2m-MLT scheme

20D is a method of combining the HARQ 2m scheme (FIG. 17B) proposed in Section 4.2.2 with a multi-layer transmission (MLT) scheme and is referred to as HARQ 2m-MLT scheme. This scheme has advantages and disadvantages of the HARQ 2m scheme except that the transmission rate is increased by multi-layer transmission. As a decoding scheme, a binding list SC decoding is used for each layer. And, we use iterative decoding between layers. Also, the effect of reducing the complexity of decoding using multiple CRCs can be obtained.

5.3.5 HARQ 3m-MLT scheme

20E is a method of combining the HARQ 3m scheme (FIG. 17C) proposed in Section 4.2.3 with a multi-layer transmission (MLT) scheme and is referred to as HARQ 3m-MLT scheme. This scheme has advantages and disadvantages of the HARQ 2m scheme except that the transmission rate is increased by multi-layer transmission. The complexity of the encoding is low. And decodes each layer using combined list SC decoding. Iterative decoding is used between layers. Also, the effect of reducing the complexity of decoding using multiple CRCs can be obtained.

6. Device Configuration

21 is a block diagram showing components of a transmitting apparatus 10 and a receiving apparatus 20 that perform the present invention.

The transmitting apparatus 10 and the receiving apparatus 20 may include RF (Radio Frequency) units 13 and 23 capable of transmitting or receiving radio signals carrying information and / or data, signals, messages, (12, 22) for storing various information related to communication, a RF unit (13, 23) and a memory (12, 22) Each comprising a processor 11, 21 configured to control the memory 12, 22 and / or the RF unit 13, 23 to perform at least one of the embodiments of the invention described above.

The memories 12 and 22 may store a program for processing and controlling the processors 11 and 21, and may temporarily store the input / output information. The memories 12 and 22 can be utilized as buffers.

Processors 11 and 21 typically control the overall operation of the various modules within the transmitting or receiving device. In particular, the processors 11 and 21 may perform various control functions to perform the present invention. The processors 11 and 21 may also be referred to as a controller, a microcontroller, a microprocessor, a microcomputer, or the like. The processors 11 and 21 may be implemented by hardware or firmware, software, or a combination thereof. When implementing the present invention using hardware, application specific integrated circuits (ASICs) or digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), FPGAs field programmable gate arrays) may be provided in the processors 400a and 400b. Meanwhile, when the present invention is implemented using firmware or software, firmware or software may be configured to include a module, a procedure, or a function for performing the functions or operations of the present invention. The firmware or software may be contained within the processors 11, 21 or may be stored in the memories 12, 22 and driven by the processors 11,

The processor 11 of the transmission apparatus 10 performs predetermined coding and modulation on signals and / or data scheduled to be transmitted from the scheduler connected to the processor 11 or the processor 11, And transmits it to the RF unit 13. For example, the processor 11 converts a data stream to be transmitted into K layers through demultiplexing, channel coding, scrambling, modulation, and the like. The encoded data stream is also referred to as a code word and is equivalent to a transport block that is a data block provided by the MAC layer. A transport block (TB) is encoded into one codeword, and each codeword is transmitted to the receiving device in the form of one or more layers. The RF unit 13 for frequency up-conversion may include an oscillator. The RF unit 13 may include Nt (where Nt is a positive integer equal to or greater than 1) transmit antennas.

The signal processing procedure of the receiving apparatus 20 is configured in reverse to the signal processing procedure of the transmitting apparatus 10. [ Under the control of the processor 21, the RF unit 23 of the receiving device 20 receives the radio signal transmitted by the transmitting device 10. The RF unit 23 may include Nr reception antennas, and the RF unit 23 performs frequency down-conversion on each of the signals received through the reception antennas to recover the baseband signals. The RF unit 23 may include an oscillator for frequency down conversion. The processor 21 may perform decoding and demodulation of the radio signal received through the reception antenna to recover data that the transmission apparatus 10 originally intended to transmit.

The RF units 13 and 23 have one or more antennas. The antenna may transmit signals processed by the RF units 13 and 23 to the outside under the control of the processors 11 and 21 or receive radio signals from the outside and transmit the signals processed by the RF unit 13 , 23). Antennas are sometimes referred to as antenna ports. Each antenna may be configured by a combination of physical antenna elements corresponding to one physical antenna or more than one physical antenna element. The signal transmitted from each antenna can not be further decomposed by the receiving apparatus 20. [ A reference signal (RS) transmitted in response to the antenna defines the antenna viewed from the perspective of the receiving apparatus 20 and indicates whether the channel is a single radio channel from one physical antenna, Enables the receiving device 20 to channel estimate for the antenna regardless of whether it is a composite channel from a plurality of physical antenna elements. That is, the antenna is defined such that a channel carrying a symbol on the antenna can be derived from the channel through which another symbol on the same antenna is transmitted. In case of an RF unit supporting a multi-input multi-output (MIMO) function for transmitting and receiving data using a plurality of antennas, it can be connected to two or more antennas.

21, the processor 11, the memory 12, and the RF unit 13 in the transmitting apparatus 10 are connected to the receiving apparatus 20 in the same way as the transmitting apparatus 10 and the receiving apparatus 20. [ And the processor 21, the memory 22 and the RF unit 23 in the transmission apparatus 20 may be configured to perform the operations of the transmission apparatus 10. [ Alternatively, the RF unit 13 of the transmission apparatus 10 and the RF unit 23 of the reception apparatus 20 may be referred to as a transceiver, respectively.

The transmitting apparatus 10 is configured to include a polar encoder according to the present invention, and the receiving apparatus 20 can be configured to include a polar decoder according to the present invention. For example, the processor 11 of the transmitting apparatus 10 may be configured to perform polar encoding according to the present invention, and the processor 21 of the receiving apparatus 20 may be configured to perform polar decoding according to the present invention. . That is, a polar encoder according to the present invention can be configured as a part of the processor 11 of the transmission apparatus 10, and a polar decoder according to the present invention can be configured as a part of the processor 21 of the reception apparatus 20 have.

In the case of an implementation by firmware or software, the method according to embodiments of the present invention may be implemented in the form of a module, a procedure, or a function for performing the functions or operations described above. For example, the software code may be stored in the memory units 12, 22 and driven by the processor 11, 21. The memory unit may be located inside or outside the processor, and may exchange data with the processor by various means already known.

The present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Accordingly, the above description should not be construed in a limiting sense in all respects and should be considered illustrative. The scope of the present invention should be determined by rational interpretation of the appended claims, and all changes within the scope of equivalents of the present invention are included in the scope of the present invention. In addition, claims that do not have an explicit citation in the claims may be combined to form an embodiment or be included in a new claim by amendment after the filing.

Embodiments of the present invention can be applied to various radio access systems. Examples of various wireless access systems include 3GPP (3rd Generation Partnership Project), 3GPP 5G NR system, or 3GPP2 system. The embodiments of the present invention can be applied not only to the various wireless access systems described above, but also to all technical fields applying the various wireless access systems.

Claims (12)

In a wireless communication system, when a terminal transmits information based on a polar code,
Generating a first packet comprising a first set of information bits, the information being encoded using a polar code;
Transmitting the first packet,
Wherein the first packet is composed of a plurality of first mini packets and a plurality of CRC (Cyclic Redundancy Check) attached to each of the plurality of first mini packets;
If an error for a specific first mini-packet transmission among the first packets is detected, encoding all or some bits of the information bit set included in the specific first mini-packet among the first information bit set using a polar code Generating a second packet comprising a second set of information bits; And
And transmitting the second packet,
Wherein the second packet comprises a plurality of CRCs (Cyclic Redundancy Check) attached to each of the one or more second mini-packets and the one or more second mini-packets,
Information transmission method. The method according to claim 1,
Wherein the second information bit set comprises:
A set of information bits included in the specific first mini packet is rearranged based on the reliability of each bit in the information bit set included in the specific first mini packet,
Information transmission method. The method according to claim 1,
The second information bit set is a set of bits in which some bits of the information bit set included in the specific first mini packet are encoded,
Wherein the bit is selected based on the reliability of each bit in the information bit set included in the specific first mini-
Information transmission method. The method according to claim 1,
Wherein the first set of information bits and the second set of information bits are encoded using a systematic polar code,
Information transmission method. The method according to claim 1,
Wherein the first set of information bits is encoded using a systematic polar code,
Wherein the second set of information bits is encoded using an unstructured polar code,
Information transmission method. 6. The method of claim 5,
Wherein the second packet comprises:
And a parity check bit equal to a parity check bit included in the first packet.
Information transmission method. The method according to claim 1,
The CRC attached to each of the plurality of first mini packets is attached for error detection for one or a plurality of first mini packets,
The CRC attached to each of the one or more second mini-packets is attached for error detection to one or more second mini-packets
Information transmission method. The method according to claim 1,
Wherein the second packet comprises:
Wherein the first set of information bits is a bit set obtained by encoding only information bits included in a first mini packet having a minimum number of CRCs attached for error checking,
Information transmission method. The method according to claim 1,
The first packet is transmitted through a plurality of layers,
Wherein the first packet comprises a CRC attached for error detection for each of the plurality of layers,
Information transmission method. In a wireless communication system, when a base station transmits information based on a polar code,
Generating a first packet comprising a first set of information bits, the information being encoded using a polar code;
Transmitting the first packet,
Wherein the first packet is composed of a plurality of first mini packets and a plurality of CRC (Cyclic Redundancy Check) attached to each of the plurality of first mini packets;
If an error for a specific first mini-packet transmission among the first packets is detected, encoding all or some bits of the information bit set included in the specific first mini-packet among the first information bit set using a polar code Generating a second packet comprising a second set of information bits; And
And transmitting the second packet,
Wherein the second packet comprises one or more second mini-packets and a plurality of CRCs attached to each of the one or more second mini-
Information transmission method. A terminal for transmitting information based on a polar code in a wireless communication system,
An encoder configured to encode the information using a polar code;
And an RF (Radio Frequency) unit configured to transmit a packet including an information bit set generated by encoding the information,
The encoder generating a first packet comprising a first set of information bits, the information being encoded using a polar code;
The RF unit transmits the first packet,
Wherein the first packet is composed of a plurality of first mini packets and a plurality of CRC (Cyclic Redundancy Check) attached to each of the plurality of first mini packets;
If an error is detected for a specific first mini-packet transmission among the first packets, the encoder converts all or some bits of the information bit set included in the specific first mini-packet among the first information bit set into a polar code Generating a second packet comprising a second set of information bits encoded using the second set of information bits; And
The RF unit transmits the second packet,
Wherein the second packet comprises one or more second mini-packets and a plurality of CRCs attached to each of the one or more second mini-
Terminal. A base station for transmitting information based on polar codes in a wireless communication system,
An encoder configured to encode the information using a polar code;
And an RF (Radio Frequency) unit configured to transmit a packet including an information bit set generated by encoding the information,
The encoder generating a first packet comprising a first set of information bits, the information being encoded using a polar code;
The RF unit transmits the first packet;
Wherein the first packet is composed of a plurality of first mini packets and a plurality of CRC (Cyclic Redundancy Check) attached to each of the plurality of first mini packets;
If an error is detected for a specific first mini-packet transmission among the first packets, the encoder converts all or some bits of the information bit set included in the specific first mini-packet among the first information bit set into a polar code Generating a second packet comprising a second set of information bits encoded using the second set of information bits; And
The RF unit transmits the second packet,
Wherein the second packet is comprised of one or more second mini-packets and a plurality of CRCs attached to each of the one or more second mini-packets.

Images