EP2057772A1

EP2057772A1 - Data transmission method using mapping on signal constellation

Info

Publication number: EP2057772A1
Application number: EP07807979A
Authority: EP
Inventors: Hyung Ho Park; Bin Chul Ihm; Min Seok Oh; Doo Hyun Sung; Jae Hoon Chung; Sung Ho Moon; Jin Soo Choi; Ki Hyoung Cho; Seung Hyun Kang
Original assignee: LG Electronics Inc
Current assignee: LG Electronics Inc
Priority date: 2006-08-07
Filing date: 2007-08-07
Publication date: 2009-05-13
Also published as: KR20080013661A; KR20080013662A; KR101253175B1; KR20080013682A; KR101287272B1; WO2008018742A1; EP2057772A4; KR101299911B1

Abstract

The data transmission method includes performing a bit remapping on signal constellation to form a plurality of data symbols, modulating the plurality of data symbols and transmitting the modulated data symbols. By the remapping, it is possible to obtain diversity gain without additional complexity.

Description

DATA TRANSMISSION METHOD USING MAPPING ON SIGNAL CONSTELLATION

Technical Field

[1] The present invention relates to a wireless communication, and more particularly, to a data transmission method which can obtain diversity gain by using mapping on signal constellation in a wireless communication system. Background Art

[2] Requirements for communication services have been rapidly increased as various multimedia services and high-quality services appear. A variety of wireless communication technologies have been studied in various technical fields so as to meet such requirements.

[3] Diversity technique is known as a technique for securing the reliability in communication system. By transmitting same data through multiple paths, data can reliably be reconstructed since data through some path can be reliable although data through the other paths are erroneous. An advantage of the diversity technique is to stably transmit and/or receive data through multiple paths independent of each other.

[4] Examples of the diversity technique include frequency diversity, time diversity and spatial diversity. The spatial diversity employs multiple transmit antennas.

[5] Since the spatial diversity is adapted on the condition that channel between transmit antenna and receive antenna is not varied at the time of transmission, inter-symbol interference (ISI) can be caused due to fast fading.

[6] Space-time code (STC) has been developed to obtain spatial diversity gain. When a transmitter uses the STC, a maximum likelihood (ML) receiver can achieve optimal performance. But it is not easy to implement the ML receiver in the actual communication system since the complexity of the ML receiver exponentially increases as number of transmit antennas and modulation order increases. A minimum means- square error (MMSE) receiver can implement easier than the ML receiver. But the MMSE receiver has to store channel information in a buffer since the STC is decoded by using cumulative combining technique. Also, the performance of the STC can be deteriorated in time- varying channel highly influenced by Doppler effect.

[7] A hybrid automatic repeat request (HARQ) technique in which forward error correction (FEC) and automatic repeat request (ARQ) are combined is known as another example for enhancing the reliability in communication system. The HARQ enhances the reliability by re-transmitting data when the data include decodable errors. An example of the HARQ is described in D. Chase, Code Combining: A maximum- likelihood decoding approach for combining an arbitrary number of noisy packets, IEEE Trans, on Commun., Vol. 33, pp. 593-607, May, 1985.

[8] In general, the HARQ can be classified into Type 1, Type II, and Type III. In Type I

HARQ, a receiver discard s data from which an error is detected and a transmitter retransmits the data. In Type II HARQ, a receiver does not discard data from which an error is detected. The receiver combines the erroneous data with re-transmitted data. The erroneous data and the re-transmitted data may have different code rates or modulation schemes. In Type III HARQ, a transmitter re-transmits data which is self- decodable.

[9] In addition, the HARQ can be classified into chase combining and IR (Incremental

Redundancy). The chase combining is a modified scheme of Type I HARQ. A receiver does not discard data from which an error is detected and combines the erroneous data with re-transmitted data. In chase combining, the erroneous data and the re-transmitted includes same information bits. In IR, a transmitter re-transmits data which incrementally includes additional redundant information.

[10] Examples of the HARQ are described in S. Lin, D. J. Costello, M. J. Miller,

Automatic repeat request error control schemes, IEEE Communications Magazine, Vol. 22, No. 12, pp. 5-17, Dec. 1984 and D. Chase, Code Combining: A maximum- likelihood decoding approach for combining an arbitrary number of noisy packets, IEEE Trans, on Commun., Vol. 33, pp. 593-607, May, 1985.

[11] Although the HARQ employs re-transmission of data, there is no spatial diversity gain when one transmit antenna is used. In multiple antenna system, the HARQ can obtain spatial diversity gain to enhance reliability of data. However, in slow fading the HARQ may have no gain since channel state is almost not varied when re-transmission occurs. In fast fading, ISI may be caused due to the spatial diversity.

[12] Therefore, there is a need for a method of enhancing the reliability of data in multiple antenna system. Disclosure of Invention Technical Problem

[13] An advantage of some aspects of the invention is that it provides a data transmission method in which remapping between data symbols is performed.

[14] Another advantage of some aspects of the invention is that it provides a data transmission method using an HARQ in which data symbol is re-transmitted after remapped on signal constellation. Technical Solution

[15] According to an aspect of the invention, there is provided a data transmission method including performing a bit remapping on signal constellation to form a plurality of data symbols, modulating the plurality of data symbols, and transmitting the modulated data symbols.

[16] According to another aspect of the invention, there is provided a data transmission method using an HARQ (Hybrid Automatic Repeat Request) including transmitting a transmission symbol, receiving a re-transmission request signal for the transmission symbol, and transmitting a re-transmission symbol obtained by remapping the transmission symbol in response to the re-transmission request signal.

[17] According to still another aspect of the invention, there is provided a transmitter including an antenna, an adaptive mapper configured to rearranges a plurality of data symbols by bits on a signal constellation, and a modulator configured to modulate the rearranged data symbols to form transmission symbols to be transmitted through the antenna.

Advantageous Effects

[18] It is possible to obtain space-time diversity gain without additional complexity. This configuration can be easily embodied because a known receiver can be used.

[19] It is possible to obtain additional diversity gain by re-transmitting symbols remapped spatially and temporally and also to enhance the performance of HARQ by minimizing the number of retransmission.

[20] It is also possible to obtain diversity gain by averagely enhancing the channel reliability of bits by remapping. By performing the bit mapping on the signal constellation every time of transmission in consideration of time and/or space multiplexing, it is possible to obtain mapping diversity. Spatial diversity gain is obtained by changing transmit antenna. Brief Description of the Drawings

[21] FIG. 1 is a block diagram illustrating a communication system according to an embodiment of the invention.

[22] FIG. 2 is a signal constellation diagram illustrating a bit error rate corresponding to a constellation mapping.

[23] FIG. 3 is a diagram illustrating an inter-symbol distance in the signal constellation diagram of FIG. 2.

[24] FIG. 4 is a diagram illustrating an example of an adaptive mapping.

[25] FIG. 5 is a diagram illustrating another example of the adaptive mapping.

[26] FIG. 6 is a diagram illustrating an example of an adaptive mapping on a space-time block code (STBC).

[27] FIG. 7 is a diagram illustrating another example of the adaptive mapping on the

STBC.

[28] FIG. 8 is a block diagram illustrating a transmitter according to an embodiment of the invention. [29] FIG. 9 is a block diagram illustrating a transmitter according to another embodiment of the invention.

[30] FIG. 10 is a diagram illustrating an example of an adaptive mapping.

[31] FIG. 11 is a diagram illustrating another example of the adaptive mapping.

[32] FIG. 12 is a diagram illustrating an example of an adaptive mapping in a system that supports multiple transmission rates. [33] FIG. 13 is a diagram illustrating another example of the adaptive mapping in the system that supports the multiple transmission rates. [34] FIG. 14 is a block diagram illustrating a transmitter according to another embodiment of the invention. [35] FIG. 15 is a block diagram illustrating a communication system according to an embodiment of the invention.

[36] FIG. 16 is a flowchart illustrating a data transmission method using the communication system shown in FIG. 15. [37] FIG. 17 is a diagram illustrating an arrangement of re-transmission symbols according to an embodiment of the invention. [38] FIG. 18 is a diagram illustrating an arrangement of re-transmission symbols according to another embodiment of the invention. [39] FIG. 19 is a diagram illustrating an arrangement of re-transmission symbols according to still another embodiment of the invention. [40] FIG. 20 is a diagram illustrating an arrangement of re-transmission symbols according to still another embodiment of the invention. [41] FIG. 21 is a diagram illustrating an arrangement of re-transmission symbols according to still another embodiment of the invention. [42] FIG. 22 is a diagram illustrating a data transmission method according to another embodiment of the invention.

[43] FIG. 23 is a graph illustrating a simulation result by SNR VS. BER (Bit Error Rate).

[44] FIG. 24 is a graph illustrating a simulation result by SNR VS. FER (Frame Error

Rate).

[45] FIG. 25 is a graph illustrating a simulation result by SNR VS. BER.

[46] FIG. 26 is a graph illustrating a simulation result by SNR VS. FER.

[47] FIG. 27 is a block diagram illustrating a transmitter according to another embodiment of the invention. [48] FIG. 28 is a diagram illustrating a re-transmission symbol in the transmitter shown in FIG. 27 [49] FIG. 29 is a block diagram illustrating a transmitter according to another embodiment of the invention. [50] FIG. 30 is a diagram illustrating a transmitter and a re-transmission symbol according to another embodiment of the invention. [51] FIG. 31 is a block diagram illustrating a communication system according to another embodiment of the invention. [52] FIG. 32 is a flowchart illustrating a hybrid automatic repeat request (HARQ) method using the communication system shown in FIG. 31. [53] FIG. 33 is a diagram illustrating an arrangement of re-transmission symbols in a 2 transmission antenna system. [54] FIG. 34 is a diagram illustrating an arrangement of re-transmission symbols in a 4 transmission antenna system. [55] FIG. 35 is a diagram illustrating an arrangement of re-transmission symbols in the 2 transmission antenna system. [56] FIG. 36 is a diagram illustrating an arrangement of re-transmission symbols in the 4 transmission antenna system. [57] FIG. 37 is a diagram illustrating an arrangement of re-transmission symbols in the 2 transmission antenna system. [58] FIG. 38 is a diagram illustrating an arrangement of re-transmission symbols in the 4 transmission antenna system. [59] FIG. 39 is a diagram illustrating an arrangement of re-transmission symbols in the 2 transmission antenna system. [60] FIG. 40 is a diagram illustrating an arrangement of re-transmission symbols in the 4 transmission antenna system. [61] FIG. 41 is a diagram illustrating an arrangement of re-transmission symbols in the 2 transmission antenna system. [62] FIG. 42 is a diagram illustrating an arrangement of re-transmission symbols in the 4 transmission antenna system. [63] FIG. 43 is a diagram illustrating an arrangement of re-transmission symbols in the 2 transmission antenna system. [64] FIG. 44 is a diagram illustrating an arrangement of re-transmission symbols in the 4 transmission antenna system. [65] FIG. 45 is a diagram illustrating a data transmission method using the HARQ according to an embodiment of the invention. [66] FIG. 46 is a block diagram illustrating a transmitter according to another embodiment of the invention.

[67] FIG. 47 is a diagram illustrating re-transmission symbols.

[68] FIG. 48 is a block diagram illustrating a transmitter using an OFDM.

[69] FIG. 49 is a block diagram illustrating another transmitter using an OFDM.

[70] FIG. 50 is a block diagram illustrating a transmitter according to another embodiment of the invention.

[71] FIG. 51 is a diagram illustrating re-transmission symbols.

[72] FIG. 52 is a flowchart illustrating a process of performing the HARQ with a chase combining scheme in a multiple-codeword multiple-antenna system.

[73] FIG. 53 is a flowchart illustrating a process of performing the HARQ with an IR scheme in the multiple-codeword multiple- antenna system.

[74] FIG. 54 is a conceptual diagram illustrating an HARQ using multiple antennas in a multiple user environment.

[75] FIG. 55 is a flowchart illustrating a data transmission method according to an embodiment of the invention.

[76] FIG. 56 is a diagram illustrating transmission blocks through multiple antennas

[77] FIG. 57 is a diagram illustrating a signal constellation in an M-QAM scheme.

[78] FIG. 58 is a diagram illustrating a BSI set in the M-QAM scheme.

[79] FIG. 59 illustrates a simulation result in which BSA schemes are compared with each other at a user equipment speed of 120 km/h.

[80] FIG. 60 illustrates a simulation result in which the BSA schemes are compared with each other at a user equipment speed of 30 km/h.

[81] FIG. 61 is a diagram illustrating an example of a BSA scheme using an arbitrary number of antennas

[82] FIG. 62 is a diagram illustrating another example of the BSA scheme using an arbitrary number of antennas and the M-QAM modulation.

[83] FIG. 63 is a diagram illustrating a mapping method of transmission blocks at the time of re-transmission in a system using a 16-QAM scheme and two transmit antennas.

[84] FIG. 64 is a graph illustrating a simulation result using the mapping method shown in FIG. 63 by SNR VS. BER.

[85] FIG. 65 is a graph illustrating a simulation result using the remapping expressed by

Equation 15 by FER Vs. SNR. Mode for the Invention

[86] The following techniques can be used in a variety of communication systems. Communications systems are widely spread to provide a variety of communication services such as voices, packets, and data. The techniques according to exemplary embodiments of the invention can be used for a downlink and an uplink. In general, the term 'downlink' means a communication directed from a base station to a user equipment and the term 'uplink' means a communication directed from a user equipment to a base station. The base station (BS) means a fixed station communicating with a user equipment, and can be called other terms such as a node-B, a BTS (Base Transceiver System), and an access point. The user equipment (UE) can be fixed or movable, and can be called other terms such as a mobile station (MS), a user terminal (UT), a subscriber station (SS), and a wireless device.

[87] A communication system may have a single transmit antenna as well as multiple transmit antennas. The communication system may be one of a multiple-input multiple-output (MIMO) system, a multiple-input single-output (MISO) system, a single-input single-output (SISO) system, and a single-input multiple-output (SIMO) system. The MIMO system uses multiple transmit antennas and multiple receiving antennas. The MISO system uses multiple transmit antennas and a single receiving antenna. The SISO system uses a single transmit antenna and a single receiving antenna. The SIMO system uses a single transmit antenna and multiple receiving antennas.

[88] The multiple access modulation system is not limited and may employ a single carrier modulation system such as a known code division multiple access (CDMA) or a multiple carrier modulation system such as OFDM (Orthogonal Frequency Division Multiplexing)/OFDMA (Orthogonal Frequency Division Multiple Access).

[89]

[90] I. MAPPING IN OPEN-LOOP TRANSMISSION DIVERSITY

[91] Here, open loop means that re-transmission is not performed and the transmission diversity system transmits at least one data symbol through a transmit antenna.

[92] FIG. 1 is a block diagram illustrating a communication system according to an exemplary embodiment of the invention.

[93] Referring to FIG. 1, a communication system includes a transmitter 10 and a receiver 20. In downlink, the transmitter 10 may be a part of a base station and the receiver 20 may be a part of a user equipment. In uplink, the transmitter 10 may be part of a user equipment and the receiver 20 may be a part of a base station. A base station may include plural receivers and plural transmitters. A user equipment may include plural receivers and plural transmitters.

[94] The transmitter 10 includes a channel encoder 11, an adaptive mapper 12, and a spatial encoder 15. The transmitter 10 includes Nt (where Nt≥l) transmit antennas 19-1, ..., 19-Nt.

[95] The channel encoder 11 receives a stream of information bits and encodes the series of information bits by the use of a predetermined coding method to form coded data. The information bits include texts, voices, images, and other data. Symbols including the coded data are hereinafter referred to as coded symbols.

[96] The adaptive mapper 12 maps the coded symbols into data symbols representing positions on a signal constellation. The data symbols are an output of the adaptive mapper 12 and can be expressed a set of bit sequences representing complex values on the signal constellation depending on the modulation scheme. The modulation scheme performed by the adaptive mapper 12 is not particularly limited and may be an m-PSK (m-Phase Shift Keying) scheme or an m-WAM (m-Quadrature Amplitude Modulation) scheme. For example, the m-PSK may be one of BPSK, QPSK, and 8-PSK. The m- QAM may be one of 16-QAM, 64-Qam, and 256-QAM. An operation of the adaptive mapper 12 will be described later in association with a data transmission method.

[97] The spatial encoder 15 allocates the data symbols output from the adaptive mapper

12 to the transmit antennas 19-1, ..., 19-Nt. The spatial encoder 15 serves to convert the serial data symbols into parallel data corresponding to the transmit antennas 19-1, ..., 19-Nt. The spatial encoder 15 processes the data symbols by the use of the space- time coding scheme and allocates the processed data symbols to the transmit antennas 19-1, ..., 19-Nt.

[98] The modulators 16-1, ..., 16-Nt modulate the parallel data by the use of a multiple access modulation scheme to form transmission symbols. One packet may include one or plural transmission symbols. The transmission symbols are transmitted through the transmit antennas 19-1, ..., 19-Nt.

[99] In the OFDM scheme, the modulators 16-1, ..., 16-Nt perform an IFFT (Inverse

Fast Fourier Transform) process. In this case, one data symbol is loaded into one subcarrier and the transmission symbols transmitted through plural carrier waves include plural data symbols. In the OFDM scheme, the transmission symbol can be said as an OFDM symbol.

[100] The receiver 20 includes a spatial decoder 22, a demapper 23, and a channel decoder 24. The receiver 20 includes Nr (where Nr≥l) receiving antennas 29-1, ..., 29-Nr.

[101] The signals received by the receiving antennas 29-1, ..., 29-Nr are demodulated by the modulators 21-1, ..., 21-Nr and are input to the spatial decoder 22. The data symbols output from the spatial decoder 22 are input to the demapper 23 and are demapped into coded data. The channel decoder 24 decodes the coded data to reconstruct the original data by the use of a predetermined decoding method.

[102] Hereinafter, S " represents a data symbol transmuted through the n-th time slot and the m-th transmit antenna. The time slot means a time interval for which one symbol is transmitted, {b b b } denotes bit sequences i to i+3 constituting the cor- i i+l i+2 bi+3 responding data symbol. However, this is only an example. The data symbol may include bit sequences representing complex values on a signal constellation and the number of bits of the data symbol may be 4 or more, or 4 or less.

[103] FIG. 2 is a signal constellation diagram illustrating a bit error rate in a constellation mapping. This signal constellation diagram is based on a gray mapping signal constellation diagram in the 16-QAM scheme. At least one bit is mapped to one symbol on the signal constellation. Examples of the constellation mapping scheme can include BPSK, QPSK, and 16-QAM and the number of bits mapped to one symbol varies depending on the types. A symbol is transmitted through an I (In-phase) channel and a Q (Quadrature) channel.

[104] Referring to FIG. 2, a signal constellation is composed of a set of two partitions.

That is, the signal constellation can be divided into first and second partitions in which the bit values at one position are equal to each other and the bit values at the other positions are different from each other. The signal constellation can be divided into two partitions of which only the first bit values from the left end are different from each other, two partitions of which only the second bit values are different from each other, two partitions of which only the third bit values are different from each other, and two partitions of which only the fourth bit values are different from each other. The sets of partitions are different depending on the bit positions, which is because inter-symbol distances are different.

[105] FIG. 3 is a diagram illustrating an inter-symbol distance in a signal constellation diagram shown in FIG. 2.

[106] Referring to FIG. 3, the inter-symbol distance varies when the first bit and the second bit are different from each other, and the inter-symbol distance does not vary when the third bit and the fourth bit are different from each other. The variation in inter-symbol distance means that the probability of the bit error at the respective bit positions varies.

[107] One symbol composed of at least one bit has an error rate varying depending on the bit positions. This fact can be utilized in a multiple antenna system. When the symbol is rearranged and transmitted in consideration of a spatial diversity gain using the multiple antennas and a time diversity gain using a time delay, it is possible to obtain an additional diversity gain. That is, by transmitting plural bits constituting one symbol through paths temporally and/or spatially different from the original path, it is possible to enhance the channel reliability of the bits constituting a data symbol as a whole.

[108] FIG. 4 is a diagram illustrating an example of an adaptive mapping scheme. Here, the number of transmit antennas is 2 (Nt=2) and the data symbols through two time slots are expressed using the 16-QAM scheme. In this case, the transmission diversity can be obtained by the mapping and the time delay. The left side shows the data symbols C ' by the use of a conventional mapping scheme and the right side shows the k data symbols S k ' by the use of an adaptive mapping scheme. C k ' and S k ' denote a data symbol transmitted through the i-th time slot in the k-th transmit antenna. [109] Referring to FIG. 4, S is constructed by swapping bit b with bit b of S and swa ^rp^rping ° bit b 4 with bit b 8 of S 2 . S 1 is constructed by ^J swa ^rp^rping ° bit b 10 with bit b 6 of

S and swapping bit b with bit b of S . S is constructed by swapping bit b with bit b of S and swapping bit b with bit b of S . S is constructed by

1 1 8 4 1 2 swapping bit b with bit b of S and swapping bit b with bit b of S . Accordingly, bits (b , b ) of S are transmitted through the same paths (through the same antennas and the same time slots) as bits (b , b ) of C at the same positions even by the adaptive mapping process, but bit b of C is shifted to S and is transmitted through a transmit antenna different from the original path. In addition, bit b of C is shifted to S and is transmitted through a transmit antenna and a time slot different from the original path. 4 bits {b , b , b , b } of C using a conventional mapping scheme are transmitted through transmit antennas and/or time slots different depending on the bits by using the adaptive mapping scheme.

[110] The adaptive mapper performs a remapping process on the bits constituting the signal constellation in different data symbols. Here, the remapping process means that the bits of the data symbols are swapped and/or replaced, but does not necessarily mean that the data symbols are mapped and a new mapping process is performed thereon. The optimal mapping method of swapping and replacing the bits of the symbols by the use of a signal constellation rearranging method can be desired depending on the number of antennas, the modulation order, the time slots, and the channel conditions.

[I l l] The number of bits temporally and spatially swapped with each other is exemplified as two per data symbol, but the number of bits swapped with each other is not limited. The bits may be swapped with each other bit by bit or by three or more bits. The swapping method is not limited to the shown example, but may be different depending on the signal constellation diagram.

[112] The number of data symbols which are transmitted for two time slots through two transmit antennas is 4 and the bits of the data symbols are spatially and/or temporally rearranged. The rearrangement means to swap and/or replace the bits with each other and may include omission and/or addition of the bits. The spatial rearrangement means to rearrange the data symbols with respect to the transmit antennas and, for example, may be to rearrange S and S . The spatial rearrangement means to rearrange the data symbols with respect to the time slots and, for example, may be to rearrange S and S 2 \

[113] The data symbols can be spatially and temporally rearranged every time slot. Alternatively, the rearrangement may be performed only once.

[114] The criterion for determining the rearrangement method is not limited, and may be different every transmission. The rearrangement of the bits of the data symbols may be different depending on the number of antennas, the modulation orders, the time slots, and the channel conditions. Alternatively, the bits may be rearranged by a fixed method regardless of the channel conditions. [115] The criterion for determining the remapping method may be determined and transmitted to the receiver by the transmitter in an open loop manner. Alternatively, the criterion may be determined and fed back to the transmitter by the receiver in a closed loop manner.

[116] When the data symbols are spatially rearranged, the signal constellation positions of the data symbols are changed. Accordingly, the signal constellation mapping is changed. That is, in the invention, the diversity is implemented by performing the signal constellation mapping spatially and temporally different on the data symbols every transmission. This is referred to as mapping diversity.

[117] The diversity gain is obtained by averagely enhancing the channel reliability of the bits of the data symbols. That is, by performing the bit mapping on the signal constellation every transmission in consideration of the spatial and temporal multiplexing, it is possible to secure the mapping diversity due to the variation in channel.

[118] FIG. 5 is a diagram illustrating another example of the adaptive mapping scheme.

[119] Referring to FIG. 5, the left side shows a conventional mapping scheme and the right side shows an adaptive mapping scheme.

O means that it is replaced with its complement. S is constructed by swapping bit b

4 9 with bit b of S and replacing it with its complement, and swapping bit b with bit b of S and replacing it with its complement. S is constructed by swapping bit b with bit b of S and replacing it with its complement, and swapping bit b with bit b of S and replacing it with its complement. S is constructed by swapping bit b with bit b of S and replacing it with its complement, and swapping bit b with bit b of S and replacing it with its complement. S is constructed by swapping bit b with bit b of S and replacing it with its complement, and swapping bit b with bit b of S and replacing it with its complement. Here, the bits spatially and temporally swapped with each other are replaced, but other bits not swapped may be replaced. [120] FIG. 6 is a diagram illustrating an example of an adaptive mapping scheme on

STBC (Space-Time Block Code). [121] Referring to FIG. 6, Alamouti STBC in a system having two transmit antennas is shown in Table 1 as well known in the art. [122] Table 1

[123] It is possible to greatly reduce the complexity of the receiver by using the Alamouti code. Here, S and S are complex conjugates of S and S . The transmission symbols S and S are temporally or spatially rearranged and the result thereof is transmitted through the STBC.

[124] The left side shows a conventional mapping scheme and the right side shows an adaptive mapping scheme. S is constructed by swapping bit b with bit b of S , swapping bit b with bit b of S , and exchanging the positions of four bits. S is constructed by swapping the bits with bits of S and exchanging the positions of four bits. That is, also in the STBC, the signal constellation mapping on the data symbols to be transmitted can be changed to obtain diversity gain.

[125] FIG. 7 is a diagram illustrating another example of the adaptive mapping scheme on the STBC.

[126] Referring to FIG. 7, the left side shows a conventional mapping scheme and the right side shows an adaptive mapping scheme. S is constructed by swapping and replacing bit b 2 with bit b 5 of S 2 , swapping and replacing bit b 3 with bit b 6 of S 2 , and exchanging the positions of four bits. S is constructed by swapping and replacing the bits with the bits of S and exchanging the positions of four bits.

[127] Here, the STBC system having two antennas has been described, but the technical spirit of the invention can be applied to a system having three or more antennas without any change. The technical spirit of the invention can be applied to a space-time trellis code (STTC) as well as the STBC system.

[128] It has been described that one data symbol is transmitted through one time slot, but plural data symbols may be transmitted through one time slot. One data symbol is modulated into one transmission symbol when one data symbol is transmitted through one time slot. In a system employing multiple subcarriers, since one data symbol is loaded into one subcarrier, plural data symbols are modulated into one transmission symbol and the transmission symbol can be said to include plural data symbols. In this case, the data symbols can be spatially and temporally rearranged and the plural data symbols transmitted through one time slot can be rearranged. In addition, the plural data symbols constituting one packet can be rearranged. Accordingly, the technical spirit of the invention includes cases where at least two data symbols different from each other are remapped by subcarriers, times, and spaces.

[129] FIG. 8 is a block diagram illustrating a transmitter according to an embodiment of the invention.

[130] Referring to FIG. 8, a transmitter 70 includes a channel encoder 71, an adaptive mapper 72, and a modulator 73. The transmitter 70 further includes one antenna 79.

[131] The adaptive mapper 72 can enhance the diversity gain by remapping the data symbol S transmitted through the first time slot Tl and the data symbol S transmitted through the second time slot T2 relative to each other. That is, it is possible to temporally remap the data symbols.

[132] In another example, plural data symbols may be transmitted through one time slot.

For example, in a packet transmitting system, plural data symbols may be included in one packet. In this case, the data symbols transmitted through one time slot can be remapped relative to each other.

[133] The adaptive mapper employing a space-time multiplexing process performs a remapping process on the bits of the data symbols to be output from the mapper using swapping and/or replacement, depending on the mapping method, the channel environment, the number of transmit antennas, and the time slots. By transmitting the remapped data symbols, it is possible to obtain an additional diversity gain, in addition to the space and time diversity gains. A receiver can perform final decoding process by swapping only the positions of soft-decision bits output from a demapper.

[134] The adaptive mapper can secure the additional diversity gain while obtaining the space and time diversity gains without any additional complexity. The adaptive mapper can be embodied without modifying the structure of a general receiver such as an MMSE receiver or a ZF (Zero-Forcing) receiver. In addition, it is possible to obtain an additional diversity gain even in a channel environment in which the space-time coding technique and the cyclic delay diversity technique may be deteriorated.

[135]

[136] II. MAPPING IN CYCLIC DELAY DEVIERSITY

[137] FIG. 9 is a block diagram illustrating a transmitter according to another embodiment of the invention. This is a transmitter using a cyclic delay diversity technique.

[138] Referring to FIG. 9, a transmitter 50 includes a channel encoder 51, an adaptive mapper 52, a spatial encoder 55, IFFT units 56-1, ..., 56-Nt, and delay units 57-1, ..., 57-(Nt-I). The transmitter 50 uses the IFFT units 56-1, ..., 56-Nt as an OFDM modulator. The transmitter 50 using the cyclic delay diversity transmits the data symbols through plural transmit antennas 59-1, ..., 59-Nt so as to have like or different power and different cyclic delays. In such a cyclic delay diversity, a signal transmitted through the plural transmit antennas is similar to a signal transmitted to a receiver through multiple paths, thereby greatly reducing the complexity in the receiver's detection of the signal.

[139] Information bits become data symbols while passing through the channel encoder

51 and the adaptive mapper 52. The data symbols pass through the spatial encoder 55 and are converted into transmission symbols in the IFFT process of the IFFT units 56-1, ..., 56-Nt. The delay units 57-1, ..., 57-(Nt-I) cyclically delay the transmission symbols. A CP (Cyclic Prefix) is inserted into the delayed transmission symbols by the CP insertion units 58-1, ..., 58-Nt) and the transmission symbols are transmitted through the transmit antennas 59-1, ..., 59-Nt.

[140] The delay times Δ , ..., Δ delayed by the delay units 57-1, ..., 57-(Nt-I) may have a constant value or may have different values by users. The delay times Δ , ..., Δ may be adjusted by receiving the corresponding information from a receiver in a feedback manner.

[141] The adaptive mapper 52 rearranges the data symbols temporally, spatially, and by subcarriers to obtain a diversity gain. The rearranged data symbols are modulated into the transmission symbols and the transmission symbols are cyclically delayed, thereby obtaining the multiplexing diversity gain.

[142] FIG. 10 is a diagram illustrating an example of an adaptive mapping. Here, one symbol S is cyclically delayed through four transmit antennas (Nt=4) and then transmitted. The data symbols transmitted through the transmit antennas are S , S (Δ ), S (Δ ), and S (Δ ). The data symbol S is not delayed, but S (Δ ) is cyclically delayed by Δ . b ^(l) denotes the i-th bit of a symbol transmitted through the k-th transmit antenna. The left side shows a data symbol C using a conventional mapping scheme and the right side shows a data symbol S using the adaptive mapping scheme. C ' and S k ' denote data symbols which is transmitted through the i-th time slot in the k-th transmit antenna. [143] Referring to FIG. 10, bits (b ⁽¹⁾, b ⁽¹⁾) of the data symbol S are transmitted through the same path (the same transmit antenna) as bits (b ⁽¹⁾, b ⁽¹⁾) equal thereto in position even by the use of the adaptive mapping, but bit b ⁽¹⁾ is shifted to S (Δ ), is cyclically

4 1 2 delayed, and is transmitted through a different transmit antenna. Bit b ⁽¹⁾ of C is shifted to S (Δ ) and is subjected to a cyclic delay and a transmit antenna different from the original path. That is, C l={b ⁽¹⁾, b ⁽¹⁾, b ⁽¹⁾, b ⁽¹⁾} using the conventional mapping is rearranged in S , S (Δ ), S (Δ ), S (Δ ) by bits using the adaptive mapping and is transmitted through transmit antennas and/or cyclic delays different from each other.

[144] The number of bits swapped in each data symbol is exemplified as two per data symbol, but the number of bits to be swapped is not limited. One bit may be swapped or three or more bits may be swapped.

[145] It has been shown that the data symbols are spatially rearranged using one time slot, but the data symbols may be temporally and spatially rearranged using two or more time slots.

[146] FIG. 11 is a diagram illustrating another example of the adaptive mapping.

[147] Referring to FIG. 11, the left side shows a conventional mapping scheme and the right side shows an adaptive mapping scheme. By spatially swapping and replacing the bits at the level of bit, it is possible to obtain an additional diversity gain.

[148] The bits spatially swapped with each other are subjected to replacement herein, but the bits not swapped may be subjected to replacement. The swapping and replacement may be performed using two or more time slots.

[149] In the examples shown in FIGs. 10 and 11, one data symbol is cyclically delayed and transmitted through the multiple transmit antennas. At this time, the transmission rate is 1. A plurality of data symbols can be transmitted through a plurality of transmit antennas. For example, four data symbols of S , S , S , and S can be cyclically delayed and transmitted through four transmit antennas. At this time, the transmission rate is 4.

[150] FIG. 12 is a diagram illustrating an example of the adaptive mapping scheme in a system supporting multiple transmission rates. Here, four data symbols S , S , S , and S are cyclically delayed and transmitted through four transmit antennas (Nt=4). The

4 data symbols transmitted through the transmit antennas become S , S (Δ ), S (Δ ), and S (Δ ). Data symbol S is not delayed and S (Δ ) is cyclically delayed by time delay Δ . The left side shows the data symbols C , C (Δ ), C (Δ ), and C (Δ ) using a con-

1 2 1 3 2 4 3 ventional mapping scheme and the right side shows the data symbols S , S (Δ ), S (Δ ), and S (Δ ) using the adaptive mapping scheme.

[151] Referring to FIG. 12, bits (b , b ) of S are transmitted through the same paths (the same antennas) as bits (b , b ) of C at the same positions even by using the adaptive mapping scheme, but bit b of C is shifted to S (Δ ), is cyclically delayed, and is transmitted throug ^σh a different transmit antenna. In addition, bit b 2 of C 1 is shifted to S4

(Δ ) and is subjected to the cyclic delay and the transmission through a transmit antenna different from the original path. That is, unlike C ={b , b , b , b }, the data symbols are rearranged in S , S (Δ ), S (Δ ), and S (Δ ) by the use of the adaptive mapping scheme and are transmitted through transmit antennas and/or cyclic delays different from each other. [152] The number of bits swapped in each data symbol is exemplified as two per data symbol, but the number of bits to be swapped is not limited. One bit may be swapped or three or more bits may be swapped. [153] It has been described that the data symbols are spatially rearranged using one time slot, but the data symbols may be temporally and spatially rearranged using two or more time slots.

[154] FIG. 13 is a diagram illustrating another example of the adaptive mapping.

[155] Referring to FIG. 13, the left side shows a conventional mapping scheme and the right side shows an adaptive mapping scheme. By spatially swapping and replacing the bits at the level of bit, it is possible to obtain an additional diversity gain. [156] The bits spatially swapped with each other are subjected to replacement herein, but the bits not swapped may be subjected to replacement. The swapping and replacement may be performed using two or more time slots. [157] The swapping and replacement by bits are only an example. The rearrangement using the swapping and/or replacement by bits of the data symbols may be performed in various methods.

[158] FIG. 14 is a block diagram illustrating a transmitter according to another embodiment of the invention. Here, the delay units of the transmitter 50 shown in FIG. 9 are moved to a frequency domain.

[159] Referring to FIG. 14, phase delay units 67-1, ..., 67-(Nt-I) are disposed between the spatial encoder 65 and the IFFT unit 68-1, ..., 68-Nt and serve to cyclically delay the phases of the symbols.

[160] The phase delay units 67-1, ..., 67-(Nt-I) in the frequency domain are equivalent to the delay unit 57-1, ..., 57-(Nt-I) in the time domain of the transmitter 50 shown in FIG. 9. The time delay in the time domain and the phase delay in the frequency domain have duality with each other.

[161] It is possible to secure the time and/or spatial diversity gain by the use of the remapped data symbols. A receiver can swap the positions relative to the soft-decision bit output from a demapper by the use of the arrangement method at the time of transmission and can perform the final decoding. Since conventional receivers such as linear receivers can be used without any change, additional complexity is not required.

[162]

[163] III. HARQ for Single Codeword

[164] FIG. 15 is a block diagram illustrating a communication system according to an embodiment of the invention.

[165] Referring to FIG. 15, a communication system includes a transmitter 100 and a receiver 200. The communication system can support an HARQ (Hybrid Automatic Repeat Request). Here, the transmitter 100 and the receiver 200 may be transceivers having both a transmitting function and a receiving function. However, in order to clarify the description of data re-transmission, one transceiver taking charge of data transmission and/or re-transmission is referred to as the transmitter 100 and the other transceiver receiving data and requesting for re-transmission is referred to as the receiver 200.

[166] The transmitter 100 includes one channel encoder 110 and one adaptive mapper 120 and thus can process one codeword at a time, which is referred to as a single codeword (SCW) scheme.

[167] The transmitter 100 includes a channel encoder 110, an adaptive mapper 120, a spatial encoder 130, a controller 150, and a receiving circuit 180. The transmitter 100 further includes Nt (where Nt≥l) transmit antennas 190-1, ..., 190-Nt.

[168] The channel encoder 110 receives and encodes a series of information bits by the use of a predetermined coding method to form coded data.

[169] The adaptive mapper 120 modulates the coded data by the use of a predetermined modulation scheme to provide data symbols. The adaptive mapper 120 maps the coded data onto data symbols representing positions on a signal constellation. The adaptive mapper 120 can adaptively map the coded data in response to a re-transmission request signal from the controller 150. The modulation scheme performed by the adaptive mapper 120 is not limited, and may be an m-PSK (m-Phase Shift Keying) scheme or an m-QAM (m-Quadrature Amplitude Modulation) scheme. The m-QAM may be one of 16-QAM, 64-QAM, and 256-QAM. An operation of the adaptive mapper 120 will be described later along with a data transmission method.

[170] The spatial encoder 130 processes the data symbols output from the adaptive mapper 120 by the use of the MIMO pre-processing method. The modulators 140-1, ..., 140-Nt modulate the symbols output from the spatial encoder 130 and transmits the modulated symbols through the transmit antennas 190-1, ..., 190-Nt. A set of symbols output from the modulators 140-1, ..., 140-Nt and transmitted during a period (or in one time slot) is referred to as a transmission symbol. In the OFDM system, the modulators 140-1, ..., 140-Nt can perform the IFFT (Inverse Fast Fourier Transform) process. In this case, one data symbol is loaded in to one subcarrier and the transmission symbol transmitted by the plural carrier waves includes the plural data symbols. In the OFDM system, the transmission symbol is an OFDM symbol.

[171] The receiving circuit 180 receives the signals transmitted from the receiver 200 through the transmit antennas (190-1, ..., 190-Nt). The receiving circuit 180 digitalizes the received signals and sends the digitalized signals to the controller 150.

[172] The controller 150 controls the entire operations of the transmitter 100. The controller 150 extracts information from the signals received by the receiving circuit 180. The operation of extracting the information includes usual demodulating and decoding operations. The extracted information can include a re-transmission request signal. The controller 150 controls the adaptive mapper 120 to prepare a retransmission symbol in response to the re-transmission request signal.

[173] The information extracted from the signals received by the receiving circuit 180 includes CQU (Channel Quality Information). The CQI may be information on a channel environment from the receiver 200 to the transmitter 100 or may be index information on the modulation and coding scheme. By the use of the CQI, the controller 150 controls the channel encoder 110 or the adaptive mapper 120 to adaptively change a coding scheme of the channel encoder 110 or a mapping scheme of the adaptive mapper 120.

[174] On the other hand, the receiver 200 includes a spatial decoder 220, a demapper

2320, a channel decoder 250, an error detector 260, a controller 270, and a transmitting circuit 280. The receiver 200 further includes Nr (where Nr≥l) receiving antennas 290-1, ..., 290-Nr. [175] The signals received by the receiving antennas 290-1, ..., 290-Nr are demodulated by the demodulators 210-1, ..., 210-Nr and are input to the spatial decoder 220. The spatial decoder 220 processes the demodulated signals by the use of an MIMO postprocessing method in response to the MIMO control signal sent from the controller 270. The MIMO control signal controls the decoding process on the basis of the space- time coding scheme of the receiver 100. The MIMO control signal can be set in advance in a memory (not shown) of the controller 270. Alternatively, the MIMO control signal may be received from the transmitter 100.

[176] The demapper 230 demaps the data symbols onto coded data in accordance with a demapping signal from the controller 270. The demapping control signal used to control the demapper 230 on the basis of the mapping scheme of the adaptive mapper 120 of the transmitter 100. The demapping control signal can be set in advance in the memory of the controller 270. Alternatively, the demapping control signal may be received from the transmitter 100.

[177] The receiver 200 includes a combiner 240 that combines the re-transmitted symbols with the previous symbols. That is, in an HARQ system with a chase combining scheme or an incremental redundancy (IR) scheme, the combiner 240 combines the previous symbols with the re-transmitted symbols. As the combining method, an equal-gain combining method of weighting the previous data and the re-transmitted data in the same way and combining them by the use of their average values may be used. Alternatively, a maximal ratio combining (MRC) method of weighting the data in the different ways may be used. The combining method is not limited and a variety of methods may be used.

[178] However, this general inventive concept is not limited to the chase combining scheme or the IR scheme, but may be applied to the HARQ system for performing the channel decoding process by the use of only the re-transmitted symbols without being combined with the previous symbols. In this case, the combiner 240 may be omitted from the receiver 200, as indicated by a dotted line in the figure.

[179] The channel decoder 250 decodes the coded data by the use of the predetermined decoding method. The error detector 260 detects whether any error exists in the decoded data bits by the use of a cyclic redundancy checking (CRC) process.

[180] The controller 270 controls the entire operations of the receiver 200 and sends the re-transmission request signal to the transmitting circuit 280. Accordingly, the controller 270 can perform a general channel encoding process, a modulation process, and the like. The controller 270 receives the detection result from the error detector 260 and determines whether the re-transmission of the symbols should be requested for. The controller 270 generates a positive acknowledgement (ACK) signal when no error is detected, and generates a negative acknowledgement (NACK) signal when an error is detected. The NACK signal may serve as a re-transmission request signal.

[181] The controller 270 measures the channel quality from the received signals and sends a CQI signal. A feedback signal such as a signal-to-noise ratio (SNR) or a bit error rate associated with the channel quality is fed back to the transmitter 100. In order to measure the channel quality, the transmission symbol transmitted from the transmitter 100 may further include a pilot symbol. The transmitting circuit 280 receives the re-transmission request signal from the controller 270 and transmits the retransmission request signal through the receiving antennas 290-1, ..., 290-Nr.

[182] FIG. 16 is a flowchart illustrating a data transmission method using the communication system shown in FIG. 15.

[183] It is assumed that the number of transmit antennas is 4 (Nt=4) and S " denotes the n-th data symbol re-transmitted through the k-th transmit antenna, {i i q q } i l+l i+2 i+3 denotes a bit sequence of i to i+3 of the corresponding data symbol. Here, i and q denote a bit of the data symbol but do not define the order or details. The data symbol may include a bit sequence representing complex values on a signal constellation and the number of bits representing the data symbol may be 4 or more bits, or 4 or less bits. The data symbol is modulated into a transmission symbol by the modulators 140-1, ..., 140-Nt and the transmission symbol is transmitted. In order to clarify the description, one transmission symbol is assumed to be obtained from one data symbol, but the transmission symbol may include a group of data symbols.

[184] Referring to FIG. 16, the transmitter 100 transmits the data symbols S , S , S , and S

(Sl 10). The data symbol S is transmitted through the first antenna 190-1, the data symbol S is transmitted through the second antenna 190-2, the data symbol S is transmitted through the third antenna 190-3, and the data symbol S is transmitted through the fourth antenna 190-4.

[185] The receiver 200 performs a channel decoding process on the received data symbols

Sl, S2, S3, and S4 and checks an error (S 120). When no error is detected, the receiver 200 transmits an ACK signal to the transmitter 100 and waits for the transmission of a next symbol. However, it is assumed that the receiver 200 detects an error and transmits a NACK signal as the re-transmission request signal (S 130).

[186] When receiving the NACK signal, the transmitter 100 transmits the re-transmission symbols S , S , S ,and S (S 140). The re-transmission symbol S is transmitted through the first antenna 190-1, the re-transmission symbol S is transmitted through the second antenna 190-2, the re-transmission symbol S is transmitted through the third antenna 190-3, and the re-transmission symbol S is transmitted through the fourth antenna 190-4. When receiving the NACK signal, the controller 150 spatially remaps the data symbols S , S , S , and S by the use of the adaptive mapper 120 to construct the re-transmission symbols S 1 , S 9 , S λ ,and S 4 . The spatial remapping [194] Referring to FIG. 17, the re-transmission symbol is formed by performing a bit remapping process on a signal constellation to the data symbols. At the first retransmission T2, the re-transmission symbols S and S are constructed by swapping

1 4 the bits (i , i ) of Sl with the bits (i , i ) of S and rearranging the bits in the symbols. The re-transmission symbols S and S are constructed by swapping the bits (q , q ) of S with the bits (q , q ) of S and rearranging the bits in the symbols. That is, the re- S , and S are constructed by spatially swapping the bits of the data symbols S , S , S , and S with each other and rearranging the bits in the

1 2 3 4 symbols.

[195] At the second re-transmission T3, the re-transmission symbols S and S are constructed by swapping the bits (q , q ) of S with the bits (q , q ) of S and rearranging the bits in the symbols. The re-transmission symbols S and S are

2 4 constructed by swapping the bits (q , q ) of S with the bits (q , q ) of S and re-

^{3 4 2 ? 8 4} 2 2 2 arranging the bits in the symbols. That is, the re-transmission symbols S , S , S , and S are constructed by spatially swapping the bits of the data symbols S , S , S , and S with each other and rearranging the bits in the symbols.

[196] The number of bits to be spatially swapped is exemplified as two, but the number of bits to be swapped is not limited. One bit may be swapped or three or more bits may be swapped.

[197] The bits of the data symbols are spatially swapped with each other at the first retransmission T2, and the bits of the data symbols are newly spatially swapped with each other and retransmitted at the second re-transmission T3. It is possible to obtain an additional diversity gain by the swapping of the bits of the data symbols.

[198] Although the second re-transmission has been described, re-transmission symbols obtained by spatially remapping the data symbols can be re-transmitted at the third or subsequent re-transmission.

[199] FIG. 18 is a diagram illustrating an arrangement of re-transmission symbols according to another embodiment of the invention.

[200] Referring to FIG. 18, the re-transmission symbols S , S , S , and S are constructed by spatially swapping the bits of the data symbols S , S , S , and S with each other and rearranging the bits in the symbols, at the first re-transmission T2.

[201] At the second re-transmission T3, the bits in the data symbols S , S , S , and S can be swapped with each other. That is, the re-transmission symbols S , S , S , and S

₂ 1 2 3 4 are constructed by swapping the bits of the data symbols S , S , S , and S with each other and replacing the LSB (Least Significant Bit) and the MSB (Most Significant Bit) with their complements. The replacement is not limited to the LSB and the MSB, but the replacement with complements may be performed independent of the LSB and the MSB. Alternatively, the intermediate bits may be replaced with the complements. and (i , i ) of the re-transmission symbol S are crossed each other. The bits (q , q ) and (i , i ) of the re-transmission symbol S are crossed each other. The retransmission symbols S and S are constructed by swapping the bits (i , i ) of S with the bits (i , i ) of S . The bits (q , q ) and (i , i ) of the re-transmission symbol S

7 8 4 ⁿ5 V 7 8 ^J 3 are crossed each other. The bits (q , q ) and (i , i ) of the re-transmission symbol S are crossed each other.

[212] The bit rearrangement of the data symbols may be performed in various methods.

The re-transmission symbols can be constructed by remapping the data symbols spatially, temporally, or by subcarriers. The re-transmission symbols may be remapped every re-transmission. Alternatively, the remapping may be performed for only one retransmission. A different remapping method or the same remapping method may be used for each remapping.

[213] The criterion for determining the remapping method is not limited. For example, the controller 150 may determined the remapping method in an open loop scheme, properly depending on the situations. The maximum Doppler frequency, the delay spread, and the like can be considered as variables for determining the remapping method. In another example, the controller 150 may receive the CQI signal and may determine the remapping method, depending on the channel quality fed back in the closed loop scheme.

[214] Since the re-transmission symbols are constructed by remapping the data symbols, such a case is called a hybrid automatic repeat request (HARQ) system of Type I or a chase combining scheme in which the entire symbols are re-transmitted. However, the technical spirit of the invention can be applied to the HARQ system of the IR scheme. That is, in the IR scheme, the entire symbols are not re-transmitted, but only redundant symbols are re-transmitted. In this case, by spatially remapping and then transmitting the redundant symbols, it is possible to secure the additional re-transmission gain.

[215] FIG. 22 is a diagram illustrating a data transmission method according to another embodiment of the invention. This is the HARQ scheme using the STBC. It is assumed that the number of transmit antennas is 2 (Nt=2) and the data symbols are S and S for the respective antennas.

[216] Referring to FIG. 22, first, the data symbol S is transmitted through the first antenna 190-1 and the data symbol S is transmitted through the second antenna 190-2.

[217] When an error is detected from the transmission symbols and thus the NACK signal is transmitted, re-transmission symbol -S is transmitted through the first antenna 190-1 and re-transmission symbol S is transmitted through the second antenna 190-2, at the first re-transmission T2.

[218] When an error is detected from the re-transmission symbols and the NACK signal is transmitted, re-transmission symbols S ' and S ' are constructed by remapping the data symbols S and S , at the second re-transmission T3. At the second re-transmission T3, the re-transmission symbol S ' is transmitted through the first antenna 190-1 and the retransmission symbol S ' is transmitted through the second antenna 190-2. [219] When an error is detected again and the NACK signal is transmitted, new retransmission symbols S ' and S ' are constructed by remapping the re-transmission symbols S and -S , at the third re-transmission T4. At the third re-transmission T4,

1 2 the re-transmission symbol -S ' is transmitted through the first antenna 190-1 and the re-transmission symbol S ' is transmitted through the second antenna 190-2.

[220] Here, the system having two antennas has been described, but the technical spirit of the invention can be applied to a system having three or more antennas without any change. The technical spirit of the invention can be applied to a space-time trellis code system as well as the STBC system.

[221] FIG. 23 is a graph illustrating a simulation result by SNR VS. BER (Bit Error Rate) and FIG. 24 is a graph illustrating a simulation result by SNR VS. FER (Frame Error Rate). For the purpose of the simulation, the chase combining scheme is used as the retransmission scheme in the 3GPP downlink and 16-QAM scheme and a 1/2 turbo code system are used. It is assumed that the number of antennas is 2 and a user speed is 100 km/h.

[222] Referring to FIGs. 23 and 24, it can be seen that the advantage of the invention is more excellent in the channel environment having high mobility than the conventional art.

[223] FIG. 25 is a graph illustrating a simulation result by SNR VS. BER and FIG. 26 is a graph illustrating a simulation result by SNR VS. FER. In FIGs. 25 and 26, it is assumed that the user speeds are 30 km/h and 150 km/h, unlike FIGs. 23 and 24.

[224] Referring to FIGs. 25 and 26, it can be seen that the advantage of the invention is more excellent than that of the conventional art and the performance is more improved with an increase in moving speed.

[225] Even in the environment having high mobility, that is, in the channel having high time selectivity, it is possible to prevent the deterioration of data by compensating for the diversity gain.

[226] FIG. 27 is a block diagram illustrating a transmitter according to an embodiment of the invention. In this case, an HARQ system using a cyclic delay diversity technique can be used.

[227] Referring to FIG. 27, a transmitter 300 includes a channel encoder 310, an adaptive mapper 320, a spatial encoder 330, a controller 350, and a receiving circuit 380. The transmitter 300 further includes IFFT units 340-1, ..., 340-Nt as the OFDM modulators.

[228] Information bits become data symbols while passing through the channel encoder 310 and the adaptive mapper 320. The data symbols pass through the spatial encoder 330 and are converted into transmission symbols by the IFFT process of the IFFT units 340-1, ..., 340-Nt. A CP (Cyclic Prefix) is inserted into the transmission symbols by the CP insertion units 345-1, ..., 345-Nt and the transmission symbols are transmitted through the transmit antennas 390-1, ..., 390-Nt. The delay units 370-1, ..., 370-(Nt-I) are disposed between the IFFT units 340-1, ..., 340-Nt and the CP insertion units 345-1, ..., 345-Nt, and cyclically delay the transmission symbols.

[229] The transmitter 300 is different from the transmitter 100 shown in FIG. 15, in that the delay units 370-1, ..., 370-(Nt-I) are added between the modulators 340-1, ..., 340-Nt and the transmit antennas 390-1, ..., 390-Nt. The other operation is equal to those of the example shown in FIG. 15.

[230] The delay units 370-1, ..., 370-(Nt-I) cyclically delay the transmission symbols transmitted through the transmit antennas 390-1, ..., 390-Nt. The delay times Dl, ..., DNt-I delayed by the delay units 370-1, ..., 370-(Nt-I) may be different depending on the users and can be adjusted by receiving the corresponding information from the receiver in a feedback manner.

[231] FIG. 28 is a diagram illustrating the re-transmission symbols used in the transmitter of FIG. 27.

[232] Referring to FIG. 28, at the first time Tl, a data symbol S is modulated into the transmission symbol and the transmission symbol is cyclically delayed and repeatedly transmitted through all the transmit antennas 390-1, ..., 390-Nt. When an error is detected from the transmitted symbols and the NACK signal is transmitted, a retransmission symbol S is constructed by remapping the data symbol S by the use of the adaptive mapper 320 at the first re-transmission T2. The re-transmission symbol S is modulated to a transmission symbol and the transmission symbol is cyclically delayed and transmitted through the transmit antennas 390-1, ..., 390-Nt.

[233] When an error is detected from the re-transmission symbol S and the NACK signal is transmitted, the re-transmission symbol S remapped by the adaptive mapper 320 is cyclically delayed and transmitted through all the transmit antennas 390-1, ..., 390-Nt at the second re-transmission T3.

[234] In another example, the delay units 370-1, ..., 370-(Nt-I) and the CP insertion units

345-1, ..., 345-Nt are replaced in position with each other. That is, the CP can be inserted after the symbols are delayed, or the symbols may be delayed after the CP is inserted.

[235] FIG. 29 is a block diagram illustrating a transmitter according to another embodiment of the invention. In the transmitter 400, the delay units of the transmitter 300 of FIG. 27 are shifted to the frequency domain.

[236] Referring to FIG. 29, phase delay units 470-1, ..., 470-(Nt-I) are disposed between the spatial encoder 430 and the IFFT units 440-1, ..., 440-Nt and serve to cyclically delay the phases of the symbols. The phase delay units 470-1, ..., 470-(Nt-I) in the frequency domain are equivalent to the delay unit 370-1, ..., 370-(Nt-I) in the time domain of the transmitter 300 shown in FIG. 27. The time delay in the time domain and the phase delay in the frequency domain have duality with each other.

[237] FIG. 30 is a diagram illustrating a transmitter and re-transmission symbols according to an embodiment of the invention.

[238] Referring to FIG. 30, a transmitter 500 includes a channel encoder 510, an adaptive mapper 520, a modulator 530, a controller 550, and a receiving circuit 560. The transmitter 500 includes one antenna 590.

[239] The data symbols output from the adaptive mapper 520 are modulated into transmission symbols by the modulator 530. Accordingly, the transmission symbols can include plural data symbols. In this case, one transmission symbol constitutes one packet. Here, three data symbols S , S , and S are included in one transmission symbol. However, this is only an example, and plural data symbols may be included in one transmission symbol, depending on the number of subcarriers.

[240] The operation of the transmitter 500 is as follows. First, three data symbols S , S , and S are transmitted at the first transmission Tl. When an error is detected from the transmitted symbols and the NACK signal is transmitted, the re-transmission symbols S , S , and S remapped by the adaptive mapper 520 are transmitted at the first retransmission T2. The re-transmission symbols can be constructed by remapping the three data symbols. When an error is detected from the re-transmission symbols and the NACK signal is transmitted, the re-transmission symbols S , S , and S 3 remapped by the adaptive mapper 520 are transmitted at the second re-transmission T3. The new re-transmission symbols can be constructed by remapping the three data symbols.

[241] The diversity is embodied by remapping the data symbols. The remapping process includes the rearrangement of the bits of two or more data symbols different from each other. This process includes the rearrangement of the data symbols transmitted through one time slot, as well as temporally and spatially.

[242] It is possible to secure an additional diversity gain while obtaining the space-time diversity gain from the corresponding channel by the use of the adaptive mapper without an increase in complexity. The invention can be applied without changing the structure of the known receiver. By transmitting the symbols temporally remapped at the time of re-transmission, it is possible to additionally secure the diversity gain, thereby minimizing the repeat request and enhancing the communication quality.

[243]

[244] IV. HARQ for Multiple Codeword [245] FIG. 31 is a block diagram illustrating a communication system according to an embodiment of the invention.

[246] Referring to FIG. 31, a communication system includes a transmitter 600 and a receiver 700. The transmitter 600 N (where N>1) channel encoders 610-1, ..., 610-N, N adaptive mappers 620-1, ..., 620-N, a spatial encoder 630, a controller 650, and a receiving circuit 680. The transmitter 650 includes Nt (where Nt≥l) transmit antennas 190-1, ..., 190-Nt. A technique of processing various codewords different in coding rate and coding scheme from each other by the use of the plural channel encoders 610-1, ..., 610-N and the adaptive mappers 620-1, ..., 620-N is called a multi codeword (MCW) technique.

[247] The channel encoders 610-1, ..., 610-N receive N information bits different from each other in parallel and encode the information bits in accordance with a predetermined coding method to form coded data. The coded data are the codewords. The coding methods applied to the information bits are independent of each other and different coding methods can be applied thereto.

[248] The adaptive mappers 620-1, ..., 620-N modulate the coded data of the information bit streams by the use of the predetermined modulation scheme to provide data symbols. The coded data are mapped to the symbols representing amplitudes and positions in a phase constellation by the adaptive mappers 620-1, ..., 620-N. The adaptive mappers 620-1, ..., 620-N can adaptively modulate the coded data in response to the re-transmission request signal from the controller 650.

[249] The spatial encoder 630 process the plural data symbols by the use of the space- time coding scheme so as to transmit the data symbols through plural transmit antennas 690-1, ..., 690-N.

[250] The modulators 640-1, ..., 640-N modulate the symbols output from the spatial encoder 130 by the use of the multiple access modulation scheme and transmit the modulated symbols through the transmit antennas 690-1, ..., 690-N. The receiving circuit 680 receives the signals from the receiver 700 through the transmit antennas 690-1, ..., 690-N. The controller 650 controls the whole operation of the transmitter 600. The controller 650 sends the re-transmission request signal transmitted from the receiving circuit 680 to the adaptive mappers 620-1, ..., 620-N so as to prepare the retransmission symbols. The information extracted from the signals received from the receiving circuit 680 includes the CQI. The controller 650 can adaptively change the coding method of the channel encoders 610-1, ..., 610-N and the mapping method of the adaptive mappers 620-1, ..., 620-N by the use of the CQI.

[251] On the other hand, the receiver 700 includes a spatial decoder 720, demappers

730-1, ..., 730-M, channel decoders 750-1, ..., 750-M, error detectors 760-1, ..., 760-M, a controller 770, and a transmitting circuit 780. The receiver 200 includes Mt (where Mt≥l) antennas 790-1, ..., 790-Mt.

[252] The signals received through the antennas 790-1, ..., 790-Mt are demodulated by the demodulators 710-1, ..., 710-M and are input to the spatial decoder 720. The spatial decoder 720 reconstructs the transmission symbols in accordance with a decoding control signal from the controller 770. The demappers 730-1, ..., 730-M demap the modulated symbols onto the coded data in accordance with a demapping control signal from the controller 70. The receiver 700 further includes combiners 740-1, ..., 740-M that combine the re-transmitted symbols with the previous symbols. As indicated by the dotted line, the combiners 740-1, ..., 740-M may be omitted from the receiver 200 that performs the channel decoding process by the use of only the retransmitted symbols without combination with the previous symbols.

[253] The channel decoders 750-1, ..., 750-M decode the coded data by the use of the predetermined decoding method. The error detectors 760-1, ..., 760-M detect whether an error exists in the decoded data bits by the CRC check. The controller 770 controls the whole operation of the receiver 700 and sends the re-transmission request signal and the like to the transmitting circuit 780. The controller 770 receives the error detection result from the error detectors 760-1, 760-M and determines whether the retransmission should be requested for. When an error is not detected, the controller 770 generates the ACK signal and when an error is detected, the controller 770 generates the NACK signal. The ACK signal or the NACK signal serves as the re-transmission request signal. The transmitting circuit 780 receives the re-transmission request signal from the controller 770 and transmits the antennas 790-1, ..., 790-M.

[254] It is assumed that the number of transmit antennas is 2 (N=2) and the data symbols are S and S for the respective antennas. In addition, in another example including four transmit antennas, it is assumed that the data symbols are S , S , S , and S . The

1 2 3 4 subscripts of the data symbols S , S , S , and S denote the transmit antenna and the superscripts denote the numbers of times of re-transmission. It is assumed for the purpose of easy description that the transmission is performed symbol by symbol, but may be performed in the unit of symbol group. Alternatively, the transmission may be performed in the unit of the whole data block or a part thereof, or in the unit of the whole data packet or a part thereof.

[255] FIG. 32 is a flowchart illustrating an HARQ method using the communication system of FIG. 31.

[256] Referring to FIG. 32, the transmitter 600 transmits S and S (, S , S ) (S610). The symbol S is transmitted through the first antenna 690-1 and the symbol S is transmitted through the second antenna 690-2. When the number of antennas is 4, the symbol S is transmitted through the third antenna 690-3 and the symbol S is transmitted through the fourth antenna 690-4. [257] The receiver 700 performs the space-time decoding process on the received symbols S and S (, S , S ) and performs the channel decoding process thereon to

1 2 3 4 check an error (S620). When no error is detected, the receiver 700 transmits the ACK signal to the transmitter 600 and waits for the transmission of next transmission symbols. However, it is assumed here that the receiver 700 detects an error and transmits the NACK signal as the re-transmission request signal (S630).

[258] When receiving the NACK signal, the transmitter 600 transmits the re-transmission symbols S and S (, S , S ) (S640). The re-transmission symbol S is transmitted

1 2 3 4 1 through the first antenna 690- 1 and the re-transmission symbol S is transmitted through the second antenna 690-2. When the number of antennas is 4, the retransmission symbol S is transmitted through the third antenna 690-3 and the retransmission symbol S is transmitted through the fourth antenna 690-4. When receiving the NACK signal, the controller 650 constructs the re-transmission symbols S and S (, S , S ) by spatially remapping the data symbols S and S (, S , S ) by the use of the adaptive mappers 620-1, ..., 620-N. Various methods can be used as the spatial remapping method used at the time of re-transmission and will be described later.

[259] The receiver 700 performs the space-time decoding process on the received retransmission symbols S and S (, S , S ) and performs the channel decoding process thereon to check an error (S650). When no error is detected, the receiver 700 transmits the ACK signal to the transmitter 600 and waits for the transmission of next transmission symbols. However, it is assumed here that the receiver 700 detects an error and transmits the NACK signal as the re-transmission request signal (S660).

[260] When receiving the NACK signal, the transmitter 600 transmits the re-transmission symbols S and S (, S , S ) (S770). The re-transmission symbol S is transmitted

1 2 3 4 1 through the first antenna 690- 1 and the re-transmission symbol S is transmitted through the second antenna 690-2. When the number of antennas is 4, the re- transmission symbol S is transmitted through the third antenna 690-3 and the retransmission symbol S is transmitted through the fourth antenna 690-4. The adaptive mappers 620-1, ..., 620-N constructs the re-transmission symbols S and S (, S , S ) by spatially remapping the data symbols S and S (, S , S ).

[261] The receiver 700 performs the space-time decoding process on the received retransmission symbols S and S and performs the channel decoding process thereon to check an error (S680).

[262] The receiver 700 transmits the ACK signal or the NACK signal to the transmitter

600 in accordance with the error detection result (S690). When the ACK signal is transmitted, the re-transmission of the corresponding symbols is ended. The repeat request using the NACK signal can be performed up to a predetermined number of times, which is n (where n≥l). When an error is detected from the n-th re-transmission, the re-transmission process is reset and the transmission of next symbols can be started. Alternatively, the transmission of the current symbols can be started again. [263] For clarity, it is assumed that the data symbols in case of two transmit antennas are

S ={a , a , a , a } and S ={b , b , b , b , b , b } and the additional transmission symbols

1 1 2 3 4 2 1 2 3 4 5 6 ^J in case of four transmit antennas are S ={c , c } and S ={d , d , d , d }. Here, a, b, c,

3 1 2 4 1 2 3 4 and d denote bit data constituting the transmission symbols and the subscript represents any bit data, but does not define the order or details. The data symbols are modulated symbols modulated by the use of the m-PSK or m-QAM. The data symbols S , S , S and S include four bit data, but they are only an example and the number of bit data included in the transmission symbol can be changed variously.

[264] The re-transmission symbols are constructed by remapping the data bits of the data symbols S and S or the data bits of the data symbols S , S , S and S . Since the

^J 1 2 ^J 1 2 3 4 transmission symbols are the modulated symbols, the remapping of the bit data of the transmission symbols can be said to be the remapping on a signal constellation. The remapping of the bit data can include the swapping or replacement of the bit data. Hereinafter, the remapping in the unit of bit is described, but the remapping may be performed in the unit of symbol in the packet transmission of transmitting plural symbols.

[265] FIG. 33 is a diagram illustrating an arrangement of the re-transmission symbols for two transmit antennas and FIG. 34 is a diagram illustrating an arrangement of the retransmission symbols for four transmit antennas. The re-transmission symbols can be constructed by swapping the bit data of the transmission symbols with each other.

[266] Referring to FIG. 33, at the first re-transmission T2 subsequent to the first transmission Tl, the re-transmission symbols S and S are constructed by swapping the bit data (a , a ) of S and the bit data (b , b ) of S with each other and rearranging the bit data in the respective symbols.

[267] Referring to FIG. 34, at the first re-transmission T2 subsequent to the first transmission Tl, the re-transmission symbols S and S are constructed by swapping the bit data (a , a ) of S and the bit data (b , b ) of S with each other and rearranging the bit data in the respective symbols. The re-transmission symbols S and S are constructed by swapping the bit data (c ) of S and the bit data (a ) of S with each other and rearranging the bit data in the respective symbols. That is, the retransmission symbols S 1 , S2 , S 3 , and S 4 are constructed by spatially swapping the bit data of the data symbols S , S , S , and S with each other and rearranging the bit data in the symbols.

2

[268] Specifically, at the second re-transmission T3, the re-transmission symbols S and

S are constructed by swapping the bit data (a ) of S and the bit data (c ) of S with each other. The re-transmission symbols S and S are constructed by swapping the bit data (b , b ) of S2 and the bit data (d , d ) of S with each other. That is, the re- spatially swapping the bit data of the data symbols S , S , S , and S with each other.

1 2 3 4

[269] The number of bit data spatially swapped with each other can be set equal to or smaller than (the number of bits/2) of the symbol having the smaller number of bits of two symbols to be swapped. This is intended to secure the minimum spatial diversity for the data transmitted by the use of the respective symbols. However, the number of bits is not limited thereto, and the number of bit data to be swapped and the order or method for the swapping is not limited.

[270] At the first re-transmission T2, the bit data of the data symbols are spatially swapped with each other and then are re-transmitted. At the second re-transmission T3, the bit data of the data symbols are newly spatially swapped with each other and then are re-transmitted. It is possible to obtain the diversity gain by swapping the bit data of the data symbols. Although the second re-transmission has been described here, the retransmission symbols which are obtained by spatially remapping the data symbols can be re-transmitted at the third or subsequent re-transmission.

[271] FIG. 35 is a diagram illustrating an arrangement of the re-transmission symbols for two transmit antennas and FIG. 36 is a diagram illustrating an arrangement of the retransmission symbols for four transmit antennas.

[272] Referring to FIG. 35, at the first re-transmission T2, the re-transmission symbols S and S are constructed by spatially swapping the bit data of the data symbols S and S and rearranging the bit data in the symbols, similarly to the re-transmission symbols at the first transmission shown in FIG. 33.

[273] At the second re-transmission T3, the bit data of the data symbols S and S can be replaced with their complements. That is, the re-transmission symbols S and S are constructed by swapping the bit data of the data symbols S and S with each other and then replacing the LSB and the MSB with their complements.

[274] Referring to FIG. 36, at the first re-transmission T2, the re-transmission symbols S

, S , S , and S are constructed by spatially swapping the bit data of the data symbols S , S , S and S with each other and rearranging the bit data in the symbols, similarly to the re-transmission symbols at the first transmission shown in FIG. 34. At the second re-transmission T3, the LSB and the MSB of the data symbols S , S , S and

1 2 3

S can be replaced with their complements.

[275] The replacement of the bit data is not limited to the LSB and the MSB, but the replacement with complements may be performed independent of the LSB and the MSB. Alternatively, the bit data of predetermined positions may be replaced with their complements. [276] Generally, the LSB and the MSB are set in the indication data such as the TFCI information and the CQI information and the LSB and the MSB can be defined in different ways, depending on the mapping schemes used in the communication system. The bit strongest against the channel variation is the defined as the MSB and the bit weakest to the channel variation is defined as the LSB. Specifically, at the initial transmission Tl, a , b , c and d are defined as the MSB, and a , b , c and d are

1 1 1 1 4 6 2 4 defined as the LSB.

[277] It is possible to obtain the diversity gain by the swapping and replacement of the bit data in the data symbols.

[278] FIG. 37 is a diagram illustrating an arrangement of the re-transmission symbols for two transmit antennas and FIG. 38 is a diagram illustrating an arrangement of the retransmission symbols for four transmit antennas.

[279] Referring to FIG. 37, at the first re-transmission T2, the re-transmission symbols S and S are constructed by spatially swapping the bit data of the data symbols S and S , rearranging the bit data, and then replacing the LSB and the MSB with their complements.

9 9

[280] At the second re-transmission T3, the re-transmission symbols S and S are constructed by spatially swapping the bit data of the data symbols S and S with each other, rearranging the bit data, and then replacing the intermediate bit data with their complements. That is, the re-transmission symbols S and S are constructed by replacing the bit data other than the replaced bit data of the first re-transmission symbols S and S with their complements.

[281] Referring to FIG. 38, at the first re-transmission T2, the re-transmission symbols S

, S , S and S are constructed by spatially swapping the bit data of the data symbols S , S , S and S with each other, rearranging the bit data, and then replacing the LSB and the MSB with their complements. At the second re-transmission T3, the re-transmission symbols S , S , S and S are constructed by spatially swapping the bit data of the data symbols S , S , S and S with each other, rearranging the bit data, and then replacing the intermediate bit data with their complements. That is, the

9 9 9 9 re-transmission symbols S , S , S and S are constructed by replacing the bit data other than the replaced bit data of the first re-transmission symbols S , S , S and S

1 2 3 4 with their complements.

[282] At the first re-transmission T2, the bit data of the data symbols are spatially swapped with each other and then are re-transmitted. At the second re-transmission T3, the bit data of the data symbols are newly spatially swapped with each other and then are re-transmitted. It is possible to obtain the diversity gain by replacing the bit data of the data symbols.

[283] FIG. 39 is a diagram illustrating an arrangement of the re-transmission symbols for two transmit antennas and FIG. 40 is a diagram illustrating an arrangement of the retransmission symbols for four transmit antennas.

[284] Referring to FIG. 39, at the first re-transmission T2, the re-transmission symbols S and S are constructed by spatially swapping the bit data of the data symbols S and S , rearranging the positions of the bit data, and then replacing the LSB and the MSB with their complements.

[285] At the second re-transmission T3, the bit data of the data symbols are swapped with each other. That is, the re-transmission symbols S and S are constructed by spatially swapping the bit data of the data symbols S and S with each other and rearranging the positions of the bit data in the symbols.

[286] Referring to FIG. 40, at the first re-transmission T2, the re-transmission symbols S

, S , S and S are constructed by spatially swapping the bit data of the data symbols S , S , S and S with each other, rearranging the positions of the bit data, and then replacing the LSB and the MSB with their complements. At the second retransmission T3, the bit data of the data symbols are swapped with each other. That is, the re-transmission symbols S , S , S and S are constructed by spatially swapping the bit data of the data symbols S , S , S and S , and rearranging the positions of the bit data in the symbols.

[287] At the first re-transmission T2, the bit data of the data symbols are spatially swapped with each other and then are re-transmitted. At the second re-transmission T3, the bit data of the data symbols are spatially swapped with each other and then are retransmitted. It is possible to obtain the diversity gain by swapping and replacing the bit data of the data symbols.

[288] FIG. 41 is a diagram illustrating an arrangement of the re-transmission symbols for two transmit antennas and FIG. 42 is a diagram illustrating an arrangement of the retransmission symbols for four transmit antennas.

[289] Referring to FIG. 41, the re-transmission symbols are constructed by swapping the bit data of the data symbols with each other. The re-transmission symbols S and S are constructed by swapping the bit data (a , a ) of S and the bit data (b , b ) of S with each other so that the bits are arranged to cross each other.

[290] Referring to FIG. 42, similarly to FIG. 41, the re-transmission symbols S and S are constructed by swapping the bit data (a , a ) of S and the bit data (b , b ) of S with each other so that the bits are arranged to cross each other. The re-transmission symbols S and S are constructed by swapping the bit data (c ) of S and the bit data (d 3 ) of S 4 with each other so that the bit data of S 3 are arrang °ed to cross each other.

[291] The swapping, replacement, and arrangement of the bit data in the symbols can be performed in various ways. [292] FIG. 43 is a diagram illustrating an arrangement of the re-transmission symbols for two transmit antennas and FIG. 44 is a diagram illustrating an arrangement of the retransmission symbols for four transmit antennas.

[293] Referring to FIGs. 43 and 44, the re-transmission symbols are constructed by swapping the bit data of the data symbols with each other. In this embodiment, only the transmission symbol for the second antenna 109-2 is re-transmitted and the transmission symbols for the other antennas are transmitted along with new data at the second transmission T2. The re-transmission is performed using a changed modulation scheme. At this time, it is preferable that the symbol from which an error is detected at the first transmission is transmitted using a modulation scheme with a low coding rate so as to enhance the transmission reliability. On the other hand, in a closed- loop multiple codeword system supporting the AMC, the controller 650 determines the MCS level (the coding rate and/or the modulation order) on the basis of the CQI information fed back from the receiver and uses a new modulation scheme for the symbols at the time of re-transmission. In an open-loop multiple codeword system, the controller can determine that all the symbols to be re-transmitted are modulated using a modulation scheme with relatively high transmission reliability.

[294] Referring to FIG. 43, the 16-QAM and the 64-QAM are applied to the symbols S and S , respectively, at the first transmission Tl, but the 64-QAM and the QPSK are applied to the symbols at the time of re-transmission T2. When a symbol modulated using the 64-QAM has been transmitted through the second antenna 690-2 but an error is detected due to the poor channel condition, the symbol is modulated and transmitted using the 16-QAM to enhance the transmission reliability at the time of retransmission. On the contrary, when the first transmission through the first antenna 690-1 is successful and the channel condition of the corresponding channel is determined as being better on the basis of feedback information, new symbol data are transmitted using the 65-QAM as a different modulation scheme at the second transmission T2. In the open loop system, the first antenna 690-1 may perform the retransmission by using the 16-QAM without any change, similarly to the first transmissi on.

[295] Referring to FIG. 44, the 16-QAM, the 64-QAM, the QPSK, and the 16-QAM are applied to the symbols S , S , S and S , respectively, at the first transmission Tl, but the 64-QAM, the 16-QAM, the 16-QAM, and the 16-QAM are applied to the symbols at the time of re-transmission T2. Here, only the data symbols through the second antenna 690-2 is re-transmitted. Specifically, the data symbols are re-transmitted using the 16-QAM with a coding rate lower than the 64-QAM. The first antenna 690-1 and the third antenna 690-3 succeeds in initial transmission and use the modulation schemes with a high coding rate at the second transmission. The fourth antenna 690-4 is maintained in modulation scheme without any change. [296] FIG. 45 is a diagram illustrating a data transmission method using the QARQ according to an embodiment of the invention. It is assumed here that the number of transmit antennas is 2 (N=2) and the data symbols for the antennas are S 1 and S2.

[297] Referring to FIG. 45, the STBC is used as the space-time coding scheme. The

Alamouti STBC is shown in Table 1. The data symbol S is transmitted through the first antenna 690- 1 and the data symbol S is transmitted through the second antenna 690-2.

[298] When an error is detected from the transmitted symbols and the NACK signal is transmitted, the re-transmission symbol -s is transmitted through the first antenna 690- 1 and the re-transmission symbol s is transmitted through the second antenna 690-2, at the first re-transmission T2.

[299] When an error is detected from the re-transmitted symbols and the NCAK signal is transmitted, the re-transmission symbols S ' and S ' are constructed by remapping the data symbols S 1 and S2. The re-transmission symbols S 1 ' and S2 ' can be constructed by swapping the bit data of the data symbols S and S with each other.

[300] At the second re-transmission T3, the re-transmission symbol S ' is transmitted through the first antenna 690- 1 and the re-transmission symbol S ' is transmitted through the second antenna 690-2.

[301] When an error is detected again and the NACK signal is transmitted, the retransmission symbols S ' and -S ' are constructed by remapping the re-transmission symbols S and -S . The re-transmission symbols S ' and -S ' can be constructed by swapping the bit data of the re-transmission symbols S and -S with each other. At the third re-transmission T4, the re-transmission symbol -s ' is transmitted through the first antenna 690-1 and the re-transmission symbol s ' is transmitted through the second antenna 690-2.

[302] Here, the system having two antennas has been described, but the technical spirit of the invention can be applied to a system having three or more antennas without any change. The technical spirit of the invention can be applied to STTC as well as STBC. In this case, a multiple codeword system is assumed. Accordingly, different coding rates and/or modulation schemes can be further used for the respective symbols, in addition to a predetermined space-time coding technique.

[303] FIG. 46 is a block diagram illustrating a transmitter according to another embodiment of the invention. FIG. 47 is a diagram illustrating the re-transmission symbols. This may be a transmitter using a cyclic delay diversity technique.

[304] Referring to FIGs. 46 and 47, the transmitter 800 includes channel encoders 810-1,

..., 810-N, adaptive mappers 820-1, ..., 820-N, a spatial encoder 830, a controller 850, and a receiving circuit 880. The transmitter 800 includes Nt (where Nt>l) antennas 890-1, ..., 890-Nt and modulators 840-1, ..., 840-Nt. [305] The embodiment of FIG. 46 is different from the embodiment of FIG. 31, in that delay units 870-1, ..., 870-(Nt-I) are additionally disposed between the modulators 840-1, ..., 840-Nt and the antennas 890-1, ..., 890-Nt. The other operations are similar to those of the embodiment of FIG. 31. The delay units 870-1, ..., 870-(Nt-I) serve to cyclically delay the data symbols to be transmitted through the antennas 890-1, ..., 890-Nt. The times (?1, ..., ?N-1) delayed by the delay units 870-1, ..., 870-(Nt-I) may have different values depending on the users and can be adjusted by the corresponding information fed back from the receiver.

[306] The operations of the transmitter 800 are as follows. First, at the first transmission

Tl, the data symbol S is cyclically delayed and repeatedly transmitted through all the antennas 890-1, ..., 890-Nt. When an error is detected from the transmitted symbol and the NACK signal is transmitted, the re-transmission symbol S remapped by the adaptive mapper 820 is cyclically delayed and transmitted through all the antennas 890-1, ..., 890-Nt at the first re-transmission T2. The re-transmission symbol S is constructed by remapping the cyclically delayed data symbol S . When an error is detected from the re-transmission symbol S and the NACK signal is transmitted, the re-transmission symbol S remapped by the adaptive mappers 820-1, ..., 820-N is cyclically delayed and transmitted through all the antennas 890-1, ..., 890-Nt at the second re-transmission T3.

[307] In the cyclic delay diversity, the spatial diversity is embodied by transmitting the same data through the plural antennas and the frequency diversity is embodied by the use of the time delay. It is possible to secure an additional re-transmission gain by remapping the re-transmission symbols.

[308] FIG. 48 is a block diagram illustrating an example of the transmitter using an

OFDM scheme. Here, the modulator of FIG. 46 is embodied in the OFDM scheme.

[309] Referring to FIG. 48, the symbols output from a spatial encoder 911 is converted into time domain samples by IFFT units 912-1, ..., 912-Nt. A CP is inserted into the samples by CP insertion units 916-1, ..., 916-Nt and the samples are transmitted through the antennas 919-1, ..., 919-Nt. Delays units 915-1, ..., 915-(Nt-I) are disposed between the IFFT units 912-1, ..., 912-Nt and the CP insertion units 916-1, ..., 916-Nt and serve to cyclically delay the samples.

[310] In another example, the positions of the delay units 915-1, ..., 915-(Nt-I) and the

CP insertion units 916-1, ..., 916-Nt can be interchanged. That is, the CP can be inserted after the symbols are delayed, but the symbols may be delayed after the CP is inserted.

[311] FIG. 49 is a block diagram illustrating another example of the transmitter using the

OFDM scheme. Here, the delay units of FIG. 46 are shifted to the frequency domain.

[312] Referring to FIG. 49, phase delay units 922-1, ..., 922-(Nt-I) are disposed between a spatial encoder 921 and IFFT units 923-1, ..., 923-Nt and serve to cyclically delay the phases of the symbols. The time delay in the time domain and the phase delay in the frequency domain have duality.

[313] FIG. 50 is a block diagram illustrating a transmitter according to an embodiment of the invention. FIG. 51 is a diagram illustrating the re-transmission symbols.

[314] Referring to FIGs. 50 and 51, the transmitter 930 includes channel encoders 931-1,

..., 931-N, adaptive mappers 932-1, ..., 932-N, a parallel-serial converter 933, a modulator 934, a controller 935, and a receiving circuit 938. The transmitter 930 includes one antenna 939.

[315] The operations of the transmitter 930 are as follows. First, the adaptive mappers

932-1, ..., 932-N outputs the plural data symbols in response to the multiple codewords output from the channel encoders 932-1, ..., 932-N. The data symbols are converted into serial signals by the parallel-serial converter 933.

[316] Here, it is assumed that three data symbols S , S and S are converted into serial

^ 1 2 3 signals. The data symbols S , S and S are transmitted at the first transmission Tl. For example, the data symbols may be transmitted at a time. In this case, a packet transmission is performed and the data symbols S , S and S constitute one packet. In another example, the plural data symbols may be sequentially transmitted through one time slot. The number of symbols to be transmitted is not limited and may be two or more. Here, since the multiple codeword system is assumed, different coding rates and/or modulation schemes may be used for the symbols. For example, the coding rate of 1/2 and the QPSK may be used for S and S , and the coding rate of 3/4 and the 16-QAM may be used for S .

[317] When an error is detected from the transmitted symbol and the NACK signal is transmitted, the re-transmission symbols S , S and S remapped by the adaptive mappers 932-1, ..., 932-N are transmitted at the first re-transmission T2. The retransmission symbols S , S and S can be constructed by swapping the bit data of the data symbols S , S and S . When an error is detected from the re-transmission symbols and the NACK signal is ransmitted, the re-transmission symbols S , S and S remapped by the adaptive mappers 932-1, ..., 932-N are transmitted at the second re-transmission T3. In this case, different coding rates and/or modulation schemes can be applied to the symbols to be re-transmitted on the basis of the feedback information.

[318] The swapping, replacement, crossing, change in modulation scheme, and delay for the re-transmission symbols can be considered when an error is detected from all the initially transmitted symbols and when an error is detected from a part of the initially transmitted symbols. When an error is detected from all the initially transmitted symbols, the bit rearrangement is performed on the symbols to be re-transmitted. However, when an error is detected from a part of the symbols, the bit rearrangement is performed on the symbols to be re-transmitted and the new symbols to be transmitted through antennas without an error.

[319] The above-mentioned re-transmission symbols may be remapped every retransmission or may be remapped only once. A different remapping method or the same remapping method may be used every remapping.

[320] The criterion for determining the remapping method is not limited. For example, the remapping method can be determined properly depending on the situations in the open- loop system. The maximum Doppler frequency, the delay spread, and the like can be considered as the variable for determining the remapping method. In another example, the remapping method can be determined depending on the channel quality fed back in the closed-loop system along with the CQI signal.

[321] FIG. 52 is a flowchart illustrating a process of performing the HARQ of the chase combining scheme in the multiple codeword multiple antenna system.

[322] Referring to FIG. 52, when an error is detected from a specific symbol of symbols

(or streams or packets) transmitted at the first transmission Tl and the NACK signal is received, the same symbols as the corresponding symbol is re-transmitted. At this time, the symbol is adaptively mapped and re-transmitted so that the data bits of the corresponding symbol are swapped, replaced, and crossed with the data bits of the other symbols.

[323] The receiver calculates the LLR values of the coded bits by the use of the demapping method which is used for the re-transmission and which is determined in advance in cooperation with the transmitter, and combines the calculated LLR values with the LLR values of the coded bits received previously, thereby enhancing the transmission reliability of data.

[324] FIG. 53 is a flowchart illustrating a procedure of performing the HARQ of the IR scheme in the multiple codeword multiple antenna system.

[325] Referring to FIG. 53, when an error is detected from a specific symbol of symbols

(or streams or packets) transmitted at the first transmission Tl, the partial IR scheme of transmitting systematic bits and punctured parity bits at the time of re-transmission may be used. Alternatively, the full IR scheme of additionally transmitting the parity bits at the time of re-transmission may be used. At this time, the transmission symbols are not re-transmitted but redundant symbols are re-transmitted. It is possible to secure the addition re-transmission gain by spatially remapping and transmitting the redundant symbols.

[326] FIG. 54 is a conceptual diagram illustrating an HARQ using multiple antennas in a multiple user environment.

[327] Referring to FIG. 54, when the symbols transmitted through one or more transmit antennas are transmitted to receiver 1 and receiver 2 and an error is detected from the

[338] The receiver detects an error from the received transmission block S (S720). When an error is not detected, the receiver transmits the ACK signal and waits for the transmission of a next transmission block. However, it is assumed here that the receiver detects an error and transmits the NACK signal as the re-transmission request signal (S730).

[339] When receiving the NACK signal, the transmitter transmits the re-transmission block S (740). When receiving the NACK signal, the adaptive mapper remaps the transmission block S by bits and/or spatially to construct the re-transmission block S . The remapping method used for the re-transmission is various, which will be described later.

[340] The receiver detects an error from the received re-transmission block Sl (S750).

[341] When not detecting an error, the receiver 200 transmits the ACK signal to the transmitter and waits for the transmission of a next symbol. However, it is assumed here that the receiver detects an error and transmits the NACK signal as the retransmission request signal (S760).

[342] When receiving the NACK signal, the transmitter transmits the remapped retransmission block S (S770). The adaptive mapper remaps the transmission block S by bits and/or spatially to construct the re-transmission block S .

[343] The receiver detects an error from the received re-transmission block S (S780). The receiver 200 transmits the ACK signal or the NACK signal to the transmitter in accordance with the error detection result (S790). When the ACK signal is transmitted, the re-transmission of the corresponding transmission block is ended. The request for re-transmission using the NACK signal can be repeated up to the n-th (where n≥l) times which is a predetermined number of times. When an error is detected due to the n-th re-transmission, the re-transmission process can be reset and the transmission of a next transmission block can be started. Alternatively, the current transmission block can be transmitted again.

[344] FIG. 56 is a diagram illustrating transmission blocks corresponding to multiple antennas.

[345] Referring to FIG. 56, TO denotes the first transmission, Tl denotes the second transmission, that is, the first re-transmission, and Tn denotes the (n+l)-th transmission, that is, the n-th re-transmission. The transmission block S is a transmission block at the first transmission, S is a transmission block at the first retransmission, and S" is a transmission block at the n-th re-transmission. [346] The optimized remapping of the bits is necessary for the re-transmission. In the data transmission method according to the invention, roughly two advantages can be obtained. First, each bit position has unique bit importance due to the characteristic of the QAM scheme. When the remapping is performed by bits, the positions on the signal constellation are changed. Accordingly, it can be said that the signal constellation mapping varies. It is called mapping diversity to embody diversity by varying the signal constellation mapping. The mapping diversity is obtained by swapping the bits of the data symbols or inverting the bits at the time of retransmission,. Second, it is possible to obtain the spatial diversity by shuffling the antennas at the time of re-transmission in the multiple antenna system. When the antennas use the same modulation scheme, the mapping diversity gain and the spatial diversity gain are independent of each other. The horizontal rearrangement scheme for obtaining the mapping diversity is referred to as a bit swapping and inversion (BSI) scheme and the vertical rearrangement scheme for obtaining the spatial diversity is referred to as a bit shuffling between antennas (BSA) scheme.

[347] The BSI scheme is first described.

[348] FIG. 57 is a diagram illustrating a signal constellation of the M-QAM scheme.

[349] Referring to FIG. 57, since the variation between antennas is not considered in the

BSI scheme, only one column of the transmission block can be considered. M signal positions exist in total and the number of index bits is Nb=log M.

[350] When the minimum distance between the signal constellation positions is D as only one of the I axis and the Q axis is viewed, position D(c) of a transmission signal c on one axis is expressed by Equation 3.

[351] MathFigure 3

[352] As only the I axis is viewed, a priori probability of the transmission signal c is A(c) and the transmission information is . d is the MSB and d is the LSB. The position of the MSB or the LSB can be

1 Nb/2 changed. For example, d may be the LSB and d may be the MSB.

1 Nb/2

[353] When a received signal is y , the LLR value of the d bit is expressed as shown.

[354] MathFigure 4 the MSB, it is possible to reduce the difference in absolute LLR value between the symbols by performing the transmission using one inversion of the first transmission. Here, only the bits relevant to the I axis are described, but the above-mentioned technique can be applied to the bits relevant to the Q axis in the same way.

[361] In the optimal BSI, the absolute value of the combined LLR received at the retransmission is constant at all the bit positions and in all the symbols. In the 16-QAM scheme and the 2x2 MIMO system, the optimal BSI can be obtained by using four BSI schemes in total of two types of bit swapping and two types of bit inversion. That is, as for the transmission block transmitted first, the transmission block having been subjected to the bit swapping is transmitted at the first re-transmission, the transmission block having been subjected to the bit inversion on the LSB is transmitted at the second re-transmission, and the transmission block having been subjected to the bit swapping and the bit inversion is transmitted at the third re-transmission. This can be expressed as shown

[362] MathFigure 5

[363]

[364]

[365]

[366] where

denotes the inversion of a bit.

[367] Table 3 shows a variation of the combined LLR values when the re-transmission is performed using the BSI set. [368] Table 3

[369] The same absolute LLR value can be obtained from the all the bit positions and all the symbols by four times of transmission. [370] The magnitude of the optimal BSI set is 4 or more in the modulation schemes of 16-QAM or greater. The difference in absolute LLR value between the bits is reduced by the bit swapping and the absolute value difference in received LLR value between the symbols is reduced by the bit inversion. Since no greatest common divisor of the absolute values of the received LLR values exist in any M-QAM scheme, bits should be transmitted through all the bit positions at the re-transmission and should be transmitted by proper inversion so as to reduce the difference between the symbols.

[371] Since (Nb/2) bit positions in total exist relative to the I axis in the M-QAM scheme and any bit should experience all the positions through the re-transmission, (Nb/2) bit swapping operations should be performed in total. The bit inversion process for reducing the absolute value difference of the received LLR between the symbols is determined on the basis of what kinds of absolute values of the received LLR values the bit positions have. Particularly, in an expression using the binary system, the necessary number of inversion operations is determined depending on the bit position having the greatest number of kinds among the bit positions. The bit located at the MSB position always has the most kinds of absolute values of the received LLR values. In the 16-QAM scheme, since the MSB has two kinds of absolute values of the received LLR values and the LSB has one kind relative to the I axis, two inversion operations are necessary. In the M-QAM scheme, since the bits at the MSB position relative to the I axis have

Mil kinds of absolute values of the received LLR values which are the most kinds, the absolute value difference of the received LLR values between the symbols can be reduced by the

{Mil inversion operations in minimum. Accordingly, the minimum number of transmissions

N for obtaining the optimal BSI in an arbitrary modulation method M-QAM can be opt expressed by Equation 6. [372] MathFigure 6

[373] FIG. 58 is a diagram illustrating a BSI set depending on the M-QAM scheme.

[374] Referring to FIG. 58, the optimal BSI set for the M-QAM scheme includes a combination of the swapping of bits using (Nb/2) cyclic shifts and

VM/2 inversions. This is true of the I axis, but is also true of the Q axis. The cyclic shift means a scheme of shifting the bits one by one and the remaining bit is disposed again at the first position. The cyclic shift is proposed because an arbitrary transmission bit should be transmitted through all the transmission bit positions. The invention is not limited to the cyclic shift, but may employ various methods. However, the arbitrary transmission bit should not be re-transmitted through the once transmitted position.

[375] The inversion process includes inverting the bits other than the leftmost bit in the transmission block formed through all the cyclic shifts bit by bit and finally inverting all the bits other than the leftmost bit. The leftmost bit is a bit that is transmitted to a position corresponding to the MSB, and the leftmost position is a bit position having the most kinds of absolute LLR values. Accordingly, the difference in absolute LLR value between the symbols for the leftmost bit is reduced by the inversion process. Since the leftmost bit circulates by the cyclic shift, the process of reducing the difference in absolute LLR value between the symbols is consequently performed on all the bits.

[376] When the re-transmission is performed the number of times smaller than the number of times of the optimum BSI, the bit swapping is first performed and then the inversion is performed. Referring to an example of Equation 5, when it is assumed that the re-transmission is performed only once, the use of the transmission block such as S or S having been subjected to the swapping of the bit positions after the first transmission is more preferable than the use of the transmission block such as S having been subjected to the process of reducing the difference in absolute LLR value swapped. Alternatively, two bits in the I axis are swapped. [388] MathFigure 10

[389] Equation 11 expresses the BSA scheme for one bit. One bit in the Q axis is swapped. Alternatively, one bit in the I axis is swapped. [390] MathFigure 11

[391] The above-mentioned combinations are all associated with the selection of the I axis and the Q axis. The bit position in the I axis or the bit position in the Q axis to be swapped can be changed.

[392] Since the I axis and the Q axis are independent of each other, the respective transmission blocks have two I-axis symbols and two Q-axis symbols. The first transmission block S and the first re-transmission block S have a time difference of several ms. Accordingly, when the speed is great, the transmission blocks may experience channel responses independent of each other. When the BSA scheme in the unit of symbols is embodied as expressed by Equation 7, each of the four symbols can experience two channel conditions.

[393] The number of channel conditions experienced by the symbols is defined as a diversity order (DO). The transmission block S expressed by Equation 7 is 2, which is equal to the DO obtained by using the same transmission matrixes every time (that is, S =S ). This means that the BSA in the unit of symbols does not provide a large advantage in performance. In the transmission block S expressed by Equation 8, two bits in the Q axis are swapped and one bit in the I axis is swapped. In this method, since two I symbols experience three channel conditions and two Q symbols experience two channel conditions, DO=(3+2)/2=2.5. In the transmission block S expressed by Equation 9, one bit in the I axis and one bit in the Q axis are swapped. In this method, the DO is 3 in maximum in consideration of one re-transmission. In the transmission block S expressed by Equation 10, only two bits in the Q axis are swapped. In this method, since all the symbols experience only two channel conditions, the DO is 2. When the speed of the user equipment decreases and thus a correlation in channel response between the re-transmissions, DO is lowered to 2 or less. In the transmission block S expressed by Equation 11, only one bit in the Q axis is swapped. In this method, since two Q symbols experience three channel conditions and two Q symbols experience two channel conditions by means of the time diversity, DO is 2.5. When the speed of the user equipment decreases, this method is lower in performance than the method expressed by Equation 8 in which three bits are swapped.

[394] FIG. 59 illustrates a simulation result in which the BSA schemes are compared with each other at a user equipment speed of 120 km/h. FIG. 60 illustrates a simulation result in which the BSA schemes are compared with each other at a user equipment speed of 30 km/h. The schemes expressed by Equations 7 to 10 are compared in the simulations and the link performance is expressed by FER (Frame Error Rate) Vs. SNR. It is checked what influence the BSA scheme has on the link performance, by considering only the bit swapping between the antennas without considering the BSI scheme.

[395] Referring to FIGs. 59 and 60, when the speed of the user equipment is great (120 km/h), the performance is enhanced with an increase in DO. When the speed of the user equipment is small (30 km/h), the temporal dependency on channel response between the re-transmissions is great and thus the scheme expressed by Equation 10 in which two bits in the Q axis are swapped is less in performance than the scheme expressed by Equation 7 in which the swapping is performed in the unit of symbols. In addition, the scheme expressed by Equation 11 in which only one bit is swapped is less in performance than the scheme expressed by quation 8 in which three bits are swapped.

[396] FIG. 61 is a diagram illustrating an example of the BSA scheme with an arbitrary number of antennas.

[397] Referring to FIG. 61, in order to allow the I symbol and the Q symbol to experience the maximum DO, it is necessary to transmit the bit information of the symbols through different antennas at the time of re-transmission. The bits can be cyclically shifted through the arbitrary number of antennas of 2 or more. In the 16-QAM scheme, when only one bit of the bits of the I symbol and the Q symbol is cyclically shifted, all the symbols can obtain the DO of 3 and thus the maximum performance can be expected.

[398] FIG. 62 is a diagram illustrating another example of the BSA scheme with an arbitrary number of antennas and an M-QAM modulation scheme.

[399] Referring to FIG. 62, the respective columns can be transmitted with different degrees of shift. In the modulation scheme higher than the 16-QAM scheme, the maximum diversity can be expected by performing a multi-step cyclic shift process. The bits of the I symbol and the Q symbol are cyclically shifted with different degrees of shift depending on the positions of the bits. For example, the degree of shift of the second bit of the I symbol may be set to 1 and the degree of shift of the third bit of the I symbol may be set to 2. The bits of one symbol can be transmitted through different antennas. Accordingly, when the number of bits (Nb/2) of the I symbol or the Q symbol is greater than the number of transmit antennas Nt, it is possible to obtain additional diversity of (Nb/2) in maximum. The degree of shift may be transmitted through the downlink or may be a value determined in advance by the transmitter and the receiver.

[400] It is considered that the re-transmission is performed two or more times. Since a relatively large time difference can exist between the re-transmissions, the channel responses between the re-transmissions are independent of each other. At this time, the previous BSA scheme does not affect the determination of the BSA scheme every retransmission. Accordingly, the BSA scheme shown in FIG. 62 can be used as the BSA scheme for the second or subsequent re-transmission without any change, regardless of the previous BSA scheme. However, when the speed of the user equipment decreases, the previous BSA and the current BSA may have some correlation. Accordingly, by changing the degree of shift of the respective columns in the multi-step cyclic shift structure depending on the number of re-transmissions, better performance can be expected.

[401] In the system using the M-QAM scheme (Nb=log M) and Nt transmit antennas, when the re-transmission is performed n times using the scheme shown in FIG. 62, the DO obtained in addition to the DO of the first transmission can be expressed by Equation 12.

[402] MathFigure 12

DO = min( AV 2 ,Nt) X n

[403] This is the maximum DO which can be obtained from the BSA scheme in the given environment. When the speed of the user equipment decreases, the DO may decrease slightly due to the correlation in channel response between the re-transmissions.

[404] FIG. 63 is a diagram illustrating a mapping method of a transmission block at the time of re-transmission in a system using the 16-QAM scheme and two transmit antennas. When it is assumed that the same QAM scheme is used for the transmit antennas, the BSI and the BSA are independent of each other and do not affect each other. Accordingly, a combination of the respective optimum methods is the whole optimum method.

[405] Referring to FIG. 63, the BSI scheme of swapping the MSB and the LSB in the first transmission block S and the BSA scheme of swapping the bits in the I channel and the Q channel bit by bit are used for the first re-transmission block S . The BSI scheme of swapping the MSB and the LSB in the first transmission block S and inverting the right bit positions in the I axis and the Q axis and the BSA scheme of swapping the bits in the I axis and the Q axis bit by bit are used for the third re-transmission block S . ⁷U ⁷1,2 ⁷1,3 ⁷1,4 #1,1 #1,2 #1,3 #1,4

12

S ⁷2,1 ⁷2,2 ⁷2,3 ⁷2,4 #2,1 #2,2 #2,3 #2,4

⁷3,1 ⁷3,2 ⁷3,3 ⁷3,4 #3,1 #3,2 #3,3 #3,4

⁷4,1 ⁷4,2 ⁷4,3 ⁷4,4 #4,1 #4,2 #4,3 #4,4

[436]

¹1.2 ⁷4,3 ⁷3,4 ⁷2,1 #1,2 #4,3 #3,4 #2,1

13

S ⁷2,2 ⁷1,3 ⁷4,4 ⁷3,1 #2,2 #1,3 #4,4 #3,1

⁷3,2 ⁷2,3 ⁷1,4 ⁷4,1 #3,2 #2,3 #1,4 #4,1

74,2 ⁷3,3 ⁷2,4 ⁷1, 1 #4,2 #3,3 #2,4 #1,1

[437]

⁷1,3 ⁷4,4 ⁷3,1 ⁷2,2 #1,3 #4,4 #3,1 #2,2

14

S ⁷2,3 ⁷1,4 ⁷4,1 ⁷3,2 #2,3 #1,4 #4,1 #3,2

⁷3,3 ⁷2,4 ⁷1,1 ⁷4,2 #3,3 #2,4 #1,1 #4,2

74,3 ⁷3,4 '2.1 ⁷1,2 #4,3 #3,4 #2,1 #1,2

[438]

⁷1,4 ⁷4,1 ⁷3,2 ⁷2,3 #1,4 #4,1 #3,2 #2,3

15

S ⁷2,4 ⁷1, 1 ⁷4,2 ⁷3,3 #2,4 #1,1 #4,2 #3,3

⁷3,4 ⁷2,1 ⁷1,2 ⁷4,3 #3,4 #2,1 #1,2 #4,3

⁷4,4 ⁷3,1 ⁷2,2 ⁷1,3 #4,4 #3,1 #2,2 #1,3

[439]

Claims

[I] A data transmission method comprising: performing a bit remapping on signal constellation to form a plurality of data symbols; modulating the plurality of data symbols; and transmitting the modulated data symbols. [2] The data transmission method according to claim 1, wherein the plurality of data symbols is remapped on time domain. [3] The data transmission method according to claim 1, wherein the plurality of data symbols is spatially remapped. [4] The data transmission method according to claim 1, wherein the plurality of data symbols is remapped on frequency domain. [5] The data transmission method according to claim 1, wherein the remapping process is to swap bits of the data symbols with each other. [6] The data transmission method according to claim 1, wherein the remapping process is to replace bits of the data symbols with each other. [7] The data transmission method according to claim 1, wherein the remapping process is performed in the unit of bits in an I axis or a Q axis on the signal constellation. [8] A data transmission method comprising: rearranging bits of a first data symbol and a second data symbol on signal constellation; modulating the rearranged first and second data symbols; transmitting the modulated first data symbol; and transmitting the modulated second data symbol. [9] The data transmission method according to claim 8, wherein the second data symbol is cyclically delayed and transmitted in comparison with the first data symbol. [10] A data transmission method using an HARQ (Hybrid Automatic Repeat Request) scheme, comprising: transmitting a transmission symbol; receiving a re-transmission request signal for the transmission symbol; and transmitting a re-transmission symbol obtained by remapping the transmission symbol in response to the re-transmission request signal.

[I I] The data transmission method according to claim 10, wherein the transmission symbol is obtained by modulating a data symbol representing a position on a signal constellation and the re-transmission symbol is obtained by remapping the data symbol.

[12] The data transmission method according to claim 10, wherein the transmission symbol and the re-transmission symbol are transmitted through one antenna.

[13] The data transmission method according to claim 10, wherein the transmission symbol and the re-transmission symbol are transmitted through a plurality of antennas.

[14] The data transmission method according to claim 13, wherein the transmission symbol is encoded with space-time block codes.

[15] The data transmission method according to claim 10, further comprising receiving again the re-transmission request signal after transmitting the retransmission symbol, and transmitting a new re-transmission symbol obtained by remapping the transmission symbol.

[16] A transmitter comprising: an antenna; an adaptive mapper configured to rearrange a plurality of data symbols by bits on a signal constellation; and a modulator configured to modulate the rearranged data symbols to form transmission symbols to be transmitted through the antenna.

[17] The transmitter according to claim 16, further comprising a delay unit configured to cyclically delay the transmission symbols.

[18] The transmitter according to claim 16, wherein a plurality of the antennas is provided and the plurality of data symbols are transmitted through the plurality of antennas, respectively.

[19] The transmitter according to claim 16, wherein the adaptive mapper remaps the plurality of data symbols to be re-transmitted in response to a request for retransmitting the transmission symbols.