EP1779573A1

EP1779573A1 - Method and apparatus for spatial channel coding/decoding in multi-channel parallel transmission

Info

Publication number: EP1779573A1
Application number: EP05773101A
Authority: EP
Inventors: Gang Philips Electronics China WU; Yueheng Philips Electronics China LI
Original assignee: Koninklijke Philips Electronics NV
Current assignee: Koninklijke Philips NV
Priority date: 2004-08-09
Filing date: 2005-08-02
Publication date: 2007-05-02
Also published as: WO2006016320A1; KR20070050436A; CN1735079A; CN1993918A; JP2008509621A; TW200607273A

Abstract

A spatial channel coding method, comprising: inputting a group of serial data to be coded according to the predefined communication rate; performing channel coding on said group of data with the relevant coding criteria according to the predefined communication mode, to output a plurality of channels of parallel coded signals, wherein there is relevant redundant information between each channel of coded signals; transmitting said channels of coded signals via a plurality of transmit antennas accordingly. This method performs coding by taking channel coding and spatial coding as a whole, so the received signals can be decoded according to the relevant redundant information between said plurality of channels of parallel signals even if there is only one receive antenna in the receiver, and thus can improve the data transmission rate of the system.

Description

METHOD AND APPARATUS FOR SPATIAL CHANNEL CODING/DECODING IN MULTI-CHANNEL PARALLEL TRANSMISSION

FIELD OF THE INVENTION

The present invention relates generally to a method and apparatus for implementing multi-channel parallel data transmission in wireless communication systems, and more particularly, to a method and apparatus for implementing multi-channel transmission in parallel by using spatial channel encoding and decoding method.

BACKGROUND ART OF THE INVENTION

With growing popularity of mobile communication, pure voice communication can no longer satisfy people's appetite for acquiring various types of information, while mobile data communication service portrays a promising picture via providing more convenient and richer contents such as SOHO, entertainment and etc. HPDS (High-speed Packet Data Service), with support of high-speed data transmission, and more particularly high-speed data service in downlink from the base station to the UE (User Equipment), has become the main objective of future wireless communication systems.

The transmission rate of data service is improved by multiplexing the frequency bands, timeslots or spreading codes in conventional single-antenna wireless communication systems. For example, in a multi-carrier system, the system allocates several frequency bands for each user, and then transmits the signals in different bands after multiplexing. While in a TDMA system, the system can also allocate different user information into different timeslots, and then transmit the signals after multiplexing. In addition, several codes could be used to transmit signals in same time slot in CDMA system. Fig.l illustrates the architecture of conventional transmitter and receiver of 3GPP UMTS FDD single-antenna system. As shown in Fig. l, after convolutional/Turbo encoding

(or namely channel coding), interleaving and symbol mapping, the data to be sent are multiplied by OVSF (Orthogonal Variable Spreading Factor) code and scrambling code successively, to get the spread and scrambled user signals. Thus, several spread user signals can be multiplexed into the same band by code division, and form RF signals through pulse shaping and RF modulation. In the receiver, after RF processing and RRC

(Root Raised Cosine) filtering and over-sampling, the RF signals received via the antenna are fed into the channel estimation unit, for estimation of characteristics of the multiple parallel propagation channels. Next, de-spreading and detecting unit first estimates channel parameters after de-spreading and then uses the estimation results to detect the received signal. After that, the detected symbol signals are processed through symbol de-mapping, de-interleaving and convolutional/Turbo decoding, to get the wanted bit signals ultimately.

In the architecture of the transmitter and receiver shown in Fig.1 , several data streams are multiplexed into the same band by code division, thus the data transmission rate can be improved.

However, currently the available radio resource in terms of frequency bands, timeslots and spreading codes are still very limited, comparing with the increasing growth and demanding requirements from wireless communication. If the data transmission rate needs further improvement, an approach is to exploit the spatial resource fully. In the recently proposed MIMO (Multiple Input Multiple Output) techniques, several transmit and receive antennas are used, to construct multiple parallel wireless channels in space domain, thus to improve the data transmission rate of the system through taking full advantage of the spatial resource. In current MIMO techniques, BLAST (Bell Labs Layered Space-Time) is a typical one to improve the data transmission rate greatly.

BLAST has various architectures, among which the BLAST architecture without any channel coding can take full advantage of the spatial channels to maximize data transmission rate since there is no redundant information in the signals to be transmitted, but unfortunately, the quality of transmitted for this architecture is very poor. To improve the transmission quality, a channel coding unit, or namely convolutional/Turbo encoder, is added into the 3GPP UMTS FDD system with the BLAST architecture described in Fig.2, playing the role same as the convolutional/Turbo encoder shown in Fig.1. Compared with Fig.l, referring to the BLAST transmitter shown in Fig.2, a single-channel data to be transmitted are fed into BLAST unit 310 after being processed by the convolutional/Turbo encoder, interleaver and symbol mapping unit; in BLAST unit 310, a single-channel data to be transmitted are divided into several independent sub-data streams; then the sub-data streams are spread and scrambled, and added with other user signals respectively to get multi-channel parallel signals; after pulse shaping and RF modulation, these parallel signals are transmitted to the receiver via multiple transmit antennas. At the receiver side, the receiver receives the signals propagated through multiple wireless channels at several receive antennas, and performs RF processing, RRC filtering and over-sampling; then, the channel estimation unit estimates the characteristics of each channel; afterwards, dispreading and detecting unit processes the received dispread signals using the channel estimation result; next, BLAST detecting unit 410 performs V-BLAST detection on the multi-channel processed signals to restore the transmitted data from each antenna, and then transforms them into a single-channel serial data. Finally, after processing of the symbol de-mapper, de-interleaver and convolutional/Turbo decoder same as those in Fig.1 , the wanted bit data are retrieved.

Referring to the transmitter and receiver in Fig.2, channel coding and BLAST are combined, thus the signal quality can be guaranteed to some extent as well when multi-channel transmission is achieved in parallel. But this BLAST architecture demodulates multi-channel signals by using the spatial characteristics of the independent spatial channels, thereby the receiver has to be equipped multiple receive antennas and the number of antennas should be more than or equal to that of the transmit antennas, and only in this way can the sub-data streams be distinguished based on the characteristics of the MIMO channels. Furthermore, it requires UE having multiple-channel RF units to process corresponding multiple-channel signals from multiple antennas. However, for a UE as the downlink HPDS receiver , it is not commercial viable considering the limitation of weight, size, battery consumption and cost. Usually, only one receive antenna is equipped for UE in current situation. Therefore, BLAST is not very suitable for providing downlink HPDS under current situations, although it can enhance data transmission rate greatly.

Besides BLAST, there are other MIMO techniques proposed for 3GPP UMTS FDD system, such as PARC (Per Antenna Rate Control), RC MPD (Rate Control Multipath Diversity) and DSTTD-SGRC (Double Space Time Transmit Diversity - Sub-Group Rate Control), and so on. But these MIMO techniques also require multiple receive antennas at the UE, and thereby not suitable for downlink high-speed transmission from the viewpoint of implementation and the cost of the UE. Based on the above analysis, MIMO techniques can achieve high-speed data transmission, but its application is restricted due to the requirement for the number of receive antennas at the UE. In prior arts, STTC (Space-Time Trellis Coding) can be adopted in single receive antenna, but existing STTC only offers the trellis diagram for two transmit antennas and 16 states at most. If more transmit antennas are needed to improve the transmission rate or more states are needed to enhance the encoding gain, it will be very complicated to design STTC in the way of plotting the trellis diagram. This restricts the application of STTC a lot. Other techniques using single receive antenna, such as space-time transmit diversity or beam shaping, only take into consideration how to improve the diversity gain or reduce the interference to improve the system performance, and thus contribute little to boost the data transmission rate.

It is, therefore, necessary to propose a method for use in multi-channel parallel data transmissions, to guarantee that the UE can achieve high data rate transmission even with only one receive antenna.

SUMMARY OF THE INVENTION

An object of the present invention is to provide a spatial channel coding and decoding method for multi-channel parallel data transmission, with which the UE can achieve multi-channel parallel data receiving even with only one receive antenna, thus to improve the data transmission rate. A method for spatial channel coding is proposed in accordance with the present invention, comprising: inputting a group of serial data to be encoded according to the predefined communication rate; performing channel coding on said group of data with the corresponding coding criteria according to the predefined communication mode, to output a plurality of channels of parallel coded signals, wherein there is relevant redundant information between each channel of the code signals; transmitting said channels of coded signals via a plurality of transmit antennas accordingly.

A method for spatial channel decoding is proposed in accordance with the present invention, comprising: receiving a plurality of channels of parallel coded signals via at least one receive antenna, wherein said plurality of channels of parallel coded signals are transmitted via a plurality of corresponding transmit antennas at a transmitting side after spatial channel coding and there is relevant redundant information between each channel of the coded signals; performing channel estimation on said plurality of wireless channels through which the coded signals are propagated according to the received pilot signal; decoding the received signals by using the channel estimation result in accordance with the spatial channel codes.

A spatial channel encoder is proposed in accordance with the present invention, comprising: a plurality of encoding branches, for receiving a group of serial data to be encoded respectively, wherein each encoding branch is composed of a plurality of registers, and performs channel coding on the group of received data by using the corresponding coding criteria according to the predefined communication mode; said plurality of encoding branches output parallel coded signals respectively, wherein there is relevant redundant information between each path of the coded signals; a plurality of transmit antennas, for transmitting said each path of coded signals accordingly.

A spatial channel decoder is proposed in accordance with the present invention, comprising: at least one receive antenna, for receiving a plurality of parallel coded signals, wherein said plurality of parallel coded signals are transmitted via a plurality of transmit antennas at the transmitting side after spatial channel coding, and there is relevant redundant information between each path of coded signals; at least one channel estimation unit, for performing channel estimation on said plurality of wireless channels propagating the coded signals according to the received pilot signal; an decoding module, for decoding the received signals by using the channel estimation result according to the spatial channel codes.

Other objects and attainments together with a fuller understanding of the invention will become apparent and appreciated by referring to the following description and claims taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

For detailed descriptions of the preferred embodiments of the invention, reference will now be made to the accompanying drawings in which like reference numerals refer to like parts, and in which: Fig.l is a block diagram illustrating the transmitter and receiver of 3GPP UMTS FDD system with single antenna;

Fig.2 is a block diagram illustrating the transmitter and receiver of 3GPP UMTS FDD system adopting BLAST technique; Fig.3 is a block diagram illustrating the transmitter and receiver adopting the spatial channel coding and decoding method proposed in accordance with preferred embodiments of the present invention, wherein the transmitter is equipped with multiple transmit antennas while the receiver has only one receive antenna;

Fig.4 shows the architecture of the spatial channel coding used in the transmitter with two transmit antennas as proposed in accordance with preferred embodiments of the present invention;

Fig.5 shows the architecture of the spatial channel coding used in the transmitter with three transmit antennas as proposed in accordance with preferred embodiments of the present invention; Fig.6 a block diagram illustrating the transmitter and receiver adopting the spatial channel coding and decoding method proposed in accordance with preferred embodiments of the present invention, wherein both the transmitter and the receiver can be equipped with a plurality of antennas;

Fig.7 shows the curves for the tested FER (Frame Error Rate) performance of the system adopting single antenna architecture with channel coding, the system adopting the proposed spatial channel coding architecture with two transmit antennas and the systems adopting two BLAST architectures, with regard to QPSK (Quadrature Phase Shift Keying) modulation;

Fig.8 shows the curves for the tested FER performance of the system adopting single antenna architecture with channel coding, the system adopting the proposed spatial channel coding architecture with two transmit antennas and the systems adopting two BLAST architectures, with regard to 8PSK modulation.

DETAILED DESCRIPTION OF THE INVENTION

In accordance with the spatial channel coding method of the present invention, channel coding is combined with the multi-channel parallel architecture, through adding some redundant information into multi-channel parallel signals so that there is spatial correlation between multi-channel parallel signals, to help the UE at the receiving side to demodulate the multi-channel parallel signals, thus the UE can receive a large mount of high-speed data traffic even with limited receive antennas. Based on the above analysis, a spatial channel coding method, which integrates channel coding and spatial encoding, is proposed in the present invention. Especifically, in the transmitter, such as base station, the number of inputted bits for one time spatial channel coding is determined first, according to the data transmission rate required by the system. Then, spatial channel coding is performed on the inputted bits respectively in the multiple branches according to the predefined coding criteria, and the multiple coded bits are transformed into multi-channel coded signals corresponding to the number of transmit antennas through modulation and mapping used by the system, so as to be transmitted to the receiving side via multiple transmit antennas. At the receiving side (UE), the receiver decodes the received signals by using the characteristics of multiple-channel obtained from channel estimation, according to the predefined coding criteria, thus multi-channel parallel coded signals are transformed into a single-channel desired data. After processing of the spatial channel coding method, among the multi-channel parallel coded signals there exist the characteristics of time correlation and space correlation, which can be exploited in decoding and thereby reduce the requirement for number of receive antenna at the UE. Detailed description will be given below to the SCC (spatial channel code) method as proposed in the present invention, in conjunction with Figs.3-5, taking an example as the UE receiver with only one receive antenna in 3GPP UMTS FDD system.

Fig.3 illustrates the architecture for the transmitter (such as base station) and the receiver (such as UE) adopting the spatial channel coding method proposed in accordance with the present invention. As shown in Fig.3, in transmitter 500, the data to be transmitted are first fed into SCC architecture 510 for spatial channel coding, which will be elaborated later in conjunction with Fig.4 and Fig.5. After being processed by SCC architecture 510, the data to be transmitted are transformed into multi-channel parallel coded signals with space correlation. Then, after being interleaved at interleaver 102, spread at OVSF spreading unit 103, scrambled at scrambler 104, overlapped at multiplexer 105, pulse shaped at pulse shaping unit 106 and modulated at RF unit 107, the multi-channel coded signals are transformed into RF signals of multiple channels and then transmitted to the wireless channel via multiple transmit antennas.

The multi-channel parallel RF signals arrive at UE receiver 600 via the wireless channel. In this embodiment, receiver 600 has only one receive antenna. The signals received by the receive antenna are the overlapped form of all signals via multiple parallel spatial channels. RF unit 208 transforms the received multi-channel RF signals into baseband signals, and sends them into RRC filtering and over-sampling unit 206, so as to transform the analog signals into discrete signals. The discrete signals obtained are processed by dispreading and detecting unit 204 and de-interleaver 202, and then fed into spatial channel decoding architecture 610. Channel estimation unit 220 estimates the characteristics of the multiple parallel spatial channels according to the received pilot signal. Then, by using the characteristics of the multiple parallel spatial channels estimated by channel estimation unit 220, spatial channel decoding architecture 610 performs the corresponding soft-decision Viterbi decoding on the de-interleaved overlapped signals, according to SCC architecture 510 employed by transmitter 500, thus gets the multi-channel parallel signals and transforms the multi-channel parallel signals into a single-channel serial data or namely the wanted data in decoding. As to the procedure of how spatial channel decoding architecture 610 performs soft-decision Viterbi decoding by using the estimated channel characteristics, it will be described hereafter. Referring to Fig.3, SCC architecture 510 is a major processing unit, responsible for channel coding and spatial coding. Two specific architectures for SCC architecture 510 are illustrated in Fig.4 and Fig.5 respectively. In Fig.4, transmitter 500 has two transmit antennas while receiver 600 has only one receive antenna. In Fig.5, transmitter 500 has three transmit antennas while receiver 600 has only one receive antenna. As shown in Fig.4 and Fig.5, in the two embodiments of the present invention, the function of symbol mapping unit 103 can also be integrated into the SCC architecture, to make the architecture more compact. In this way, after processing the data bits inputted into the SCC architecture, multiple channels of phase/amplitude modulated symbols can be obtained. Further, the symbol mapping unit can also be placed outside the SCC architecture. The SCC architecture will be described below, referring to Fig.4 where transmitter

500 has two transmit antennas while receiver 600 has only one receive antenna. As Fig.4 shows, the SCC architecture comprises 4 parallel branches of shift registers, each of which is represented by a group of registers D, and two QPSK mapping units, wherein each branch of shift register implements a channel coding, and thus is also called as channel coding branch. In this embodiment, assuming that the system requires the transmitter with two transmit antennas shown in Fig.3 to have data transmission rate at 2b/s/Hz (b/s/Hz denotes the number of bits rate per unit spectrum and unit time) and QPSK modulation is adopted, the number of bits inputted into the SCC architecture for one-time encoding is 2 according to the data transmission rate 2b/s/Hz, based on the essential idea of the present invention. Meanwhile, each transmit antenna adopts QPSK modulation where two bits correspond to one symbol and the transmitter has two transmit antennas, which means the number of the outputted coded bits is 4, so the SCC architecture has 4 channel coding branches and the coded bits outputted by two branches are used to generate one QPSK symbol. Thus, the coding rate of the SCC architecture is 1/2. Once the number of channel coding branches is determined, the number of registers for channel coding in each channel coding branch can be set to 9, i.e. the encoder of each branch in Fig.4 has 9 registers D, to improve the anti-interfering capability of the signals. Two input bit information is needed for coding once, so only 7 of the 9 registers D maintain the shift states after the last coding, that is, each channel coding branch has 2⁷=128 states totally. Noticeably, the number of the states of the channel coding branch can be set according to the requirement of transmission quality for practical systems. If the system has a high requirement for signal quality, more registers are needed.

According to the above setting, the SCC architecture with data transmission rate as 2b/s/Hz and 128 states, is constructed for use situation that there are two transmit antennas and one receive antenna, which are illustrated in Fig.4.

Referring to Fig.4, two serial data bits bl and b2 are shifted to 4 parallel encoding branches at same time. Each processing branch performs coding according to the pre-designed coding criteria, for example, the generation code GO for the first channel coding branch is 101110011 and the generation code Gl for the second channel coding branch is 010011100, wherein each bit of the generation code is corresponding to a register.

For each branch, the specific coding procedure is that the coded bit information is obtained by performing modulo-2 addition on all bit states stored in the registers whose corresponding generation code bit is set as 1. Then, QPSK mapping is performed by combining the coded bits of the first branch and those of the second branch, thus to get the symbol ci to be transmitted via the first transmit antenna. Similarly, QPSK mapping is performed on the coded bits of the third and fourth branches, to get the symbol C2 to be transmitted via the second transmit antenna. In this way, a single-channel serial data are coded and mapped into two channels of parallel signals coded according to the architecture shown in Fig.4, and space correlation exists between the two channels of parallel signals.

Referring to the SCC architecture in Fig.4, a suitable coding criteria needs to be selected to improve the anti-interference capability of the signals and minimize the BER

(Bit Error Rate) of the signals. Accordingly, the method is proposed to make a system design on the spatial channel codes. Since a basis of the coding system design is to be able to construct codes generation matrix G, the architecture in Fig.4 will be generalized below to construct the mathematic model for spatial channel codes. First, assuming that the data information bits inputted into the SCC architecture are

B = [b, ,..., b_κ ] , wherein K denotes the bit number of encoding SCC once. According to the idea of the present invention, K is decided by the data transmission rate the system requires, for example, K=2 when the data transmission rate is 2b/s/Hz in the architecture of Fig.4. Thus, the symbols C_n corresponding to the nth transmit antenna wherein is the number of transmit antennas, can be represented as follows: c_n = Φ[(D - G^_{.ρ+/ ρ}J] (1) where D = [B M] , M = [b_κ+i ,..., b_κ+s ] represents the information bits in current states in the registers, S is the number of the registers for saving the state information according to the system performance requirement, for example, S =7 in the architecture shown in Fig.4, that is, D=[ bi, b₂, ,bκ bκ+ι, b_κ+2 ,bκ+s]- G is codes generation matrix, where the superscript T of codes generation matrix G denotes the transpose of the matrix; Qn-Q+l:Qn in the subscript denotes the values from Qn-Q+1 row to Qn row are used to calculate c_n. Q in equation (1) represents the modulation used by the system, for example, Q = 2 if the system adopts QPSK modulation and Q = 4 if the system adopts 16QAM (Quadrature Amplitude Modulation) modulation. From the equation, it can be seen that the number of a group of channel coding branches corresponding to the symbols C_n transmitted by each transmit antenna is decided by the value of Q corresponding to the modulation mode. For example, if the architecture shown in Fig.4 adopts QPSK, i.e. Q =2, the generation codes for generating C₁ are the first row and second row of the codes generation matrix, i.e. the first and second channel coding branches as shown in Fig.4. Each row in G corresponds to a generation codes of one channel coding branch, that is, each element in codes generation matrix G corresponds to a register D in the shift registers, and the bit in the register D will be extracted to perform modulo-2 addition if the element is l.Φ(-) of equation (1) represents the modulation procedure for mapping the coded value into phase symbol. Referring to the architecture shown in Fig.4, the system uses QPSK modulation, then Φ( ) is represented as Φ(x)=exp(πry x / 2), x= 1 ,2....4.

As described above, the spatial channel coding procedure can be represented by equation (1), and thus the mathematical model for spatial channel coding is constructed, wherein the number of rows and that of columns of the codes generation matrix are determined according to the actual data transmission rate, the number of transmit antennas, and the modulation method used by the system. The next step described below will focus on how to decide the elements of the codes generation matrix G, i.e. search the suitable codes generation matrix to obtain the optimum coding effect.

If the product of the number of transmit antennas and that of receive antennas is not more than 3, the following criteria can be used to search codes generation matrix G: in the trellis diagram generated according to the codes generation matrix G, the difference matrix

B, being constructed with the output symbols corresponding to each decoding path, can maximize the minimum of the rank/product of the matrix A=B*B' wherein B' represents the conjugate transpose of the difference matrix B. In this embodiment, the number of transmit antennas N=2 and the number of receive antennas M=I, so the criteria can be used to search codes generation matrix G In the searching procedure, the trellis diagram corresponding to each codes generation matrix G is judged by the criteria of maximizing the minimum of the rank/product of the matrix A and thus the optimum codes generation matrix G as represented by equation (2) can be found:

1 0 1 1 1 0 0 1 I

0 1 0 0 1 1 1 0 0

When N = 2, M = 1, G (2)

0 1 0 0 0 1 0 1 1

1 0 0 1 0 0 0 0 1 The implementation architecture of the codes generation matrix G is shown in Fig.4.

Substituting the codes generation matrix G in equation (2) into equation (1) , then we can get the parallel coded signals corresponding to the two transmit antennas, i.e. the output signals of the architecture shown in Fig.4. Referring to the architecture of transmitter 500 shown in Fig.3, the two channels of parallel coded signals are transmitted to the UE receiver via the two transmit antennas after being interleaved, spread, scrambled, pulse shaped and modulated. In the receiver, the spatial channel decoding architecture 610 shown in Fig.3 performs soft-decision Viterbi decoding on the received signals according to the trellis diagram generated from the codes generation matrix G in equation (2), based on the channel characteristics estimated by the channel estimation unit 220.

Soft-decision Viterbi decoding works as follows: the spatial channel decoding architecture searches each path in the trellis diagram determined by the number of states of the coding architecture according to the code generation matrix G and finds the optimum path with minimum error of the received signals, thus the inputted data bits corresponding to the optimum path are the wanted data. In the embodiment of the present invention, receiver 600 has only one receive antenna, the received signals are the sum of two channels of parallel transmitted signals, for example, r = h_λ x C_x + h₂ x C₁ + n , wherein h is channel characteristics and n is Gauss noise. Thus in path search procedure, spatial channel decoding architecture 610 takes the estimated channel characteristics h_/ and fi₂ as the weights to multiply the coded signals and c₂ on the search path and then add them, to get the signals to be judged, i.e. r'= A₁ x c\ +h₂ x c'₂ . Then, spatial channel decoding architecture 610 compares the signals to be judged r' with the received signals r, to generate a decision metric. The decision metric reflects the difference between r' and the actually received signals r. In searching procedure, soft-decision Viterbi decoding only keeps the path with the minimum decision metric as the optimum path, thus to retrieve the wanted user data.

The above description illustrated the proposed spatial channel coding method and architecture, taking an example as the transmitter with two transmit antennas and the receiver with one receive antenna The method proposed in the present invention is also suitable for the case where the transmitter has three transmit antennas (N=3) while the receiver has only one receive antenna (M=I). In this case, assuming that the data transmission rate required by the system is 3b/s/Hz, K=3, that is, the number of bits inputted for one time encoding is 3, and if 8PSK modulation is used by the system, Q=3 and 8PSK modulation can be represented as Φ(x)=exp(π j x / 8), x=l,2....8. According to the requirement of equation (1), the number of channel coding branches corresponding to each transmit antenna is 3. If at this time the system performance requirement can be met by adopting the channel coding branch with 128 states (S=7), the codes generation matrix with N=3, M=I and S=7 has 9 rows and 10 columns. Meanwhile, the suitable codes generation matrix G can be found with the criteria of maximizing the minimum of the rank/product of the matrix A. After searching, the optimum codes generation matrix G with N=3, M=I and S=7 can be expressed by equation (3) as follows:

When N = 3,Λ/ = 1, G = 11 11 11 00 00 00 0 1

Three successively inputted data bits bl, b2 and b3 are coded in accordance with the codes generation matrix G in equation (3). According to the requirement for the number of rows of the codes generation matrix for generating C_n being Qn-Q+]. Qn in equation (1), the inputted symbols Ci are the mapped symbols of the coded bits generated from the first three rows in the codes generation matrix G.

Fig.5 illustrates the spatial channel coding architecture with 3b/s/Hz and 128 states for use in three transmit antennas and one receive antenna is constructed in accordance with equation (3). The basic principle is same as that in Fig.4, with difference in that 3 input bits can be processed in one time decoding once and there are 9 parallel channel coding branches. The coded bits outputted by every three of the 9 branches are combined to make 8PSK mapping and thus to get three channels of parallel coded signals. By utilizing the spatial channel coding architecture shown in Fig.5, the transmitter correspondingly has three channels of interleaving, spreading, scrambling, pulse shaping and RF modulating units, so as to transform the three paths of coded signals into RF signals to be transmitted via the corresponding transmit antennas.

At the receiving side, since single-antenna receiver is adopted again, the architecture of the receiver is completely the same as that in Fig.3. But the received signals comprise signals from three channels, i.e. r - h_x x c, + A₂ x c₂ + A₃ x C₃ + n , so the channel estimation unit of the receiver in Fig.3 needs to estimate the channel characteristics hi, h₂ and hi of the three channels. When the spatial channel decoding architecture performs soft-decision Viterbi decoding, the signals to be judged are the sum of three weighted coded signals, that isr'= A₁ χ c\ +h₂ χ c'₂ +A₃ χ c'₃. The wanted user data can be retrieved through using the decision metric generated by comparing r' with the received signals r to select the optimum path in the trellis diagram.

The above description goes to the case where transmitter 500 has two or three transmit antennas while receiver 600 has only one receive antenna, in conjunction with

Fig.4 and Fig.5. Further, the proposed method is not limited to the two cases, and can be applied in cases where there are more transmit antennas, and also be applied in cases where the receiver has several receive antennas.

Fig.6 illustrates the architecture for the transmitter equipped with multiple transmit antennas, and that for the receiver with multiple receive antennas in accordance with the present invention. Compared with Fig.3, receiver 700 has multiple receive antennas in Fig.6, and accordingly the receiver includes multiple receive processing branches, each of which has the same structure with the receive processing branch of the single-receive antenna shown in Fig.3, comprising RF unit 208, RRC filtering and over-sampling unit 206, dispreading and detecting unit 204, de-interleaver 202 and channel estimation unit 220.

The received signals processed by each receive processing branch, are sent to spatial channel decoding architecture 710, along with the channel characteristics estimated by the channel estimation unit in each branch, for decoding. Different from the case for single receive antenna, when performing soft-decision Viterbi decoding, spatial channel decoding architecture 710 can weight and then add the received multi-channel signals to get the optimum received signals, decode the received signals by using the channel characteristics estimated from channel estimation, and retrieve the wanted user data ultimately. It can be seen here that the receiver can employ multiple antenna reception to improve the receive diversity gain of signals, thus to decrease BER of signals. Therefore, when the receiver has multiple receive antennas, the spatial channel coding rate can be increased and the data transmission rate can be further improved. Referring to Fig.6, the method proposed in the present invention, can be applied for single receive antenna, and multiple receive antennas as well. It's to be noticed that the above judging criteria for finding the optimum codes generation matrix will change when the product of the number of transmit antennas and that of receive antennas is more than 3. Researches indicate that the multiple channels of multipath fading channels will turn into

AWGN channels when there are more than three diversity branches. Thus, the judging criteria for coding spatial channel codes will become the same as that for conventional TCM, for example, to guarantee that the minimum Euclidean distance can be maximized. Even when there are more antennas, TCM coding can be chosen directly as the spatial channel coding architecture, thus TCM, used to be applicable only for single antenna, can be applied for multiple antennas hereafter.

. From the descriptions to the embodiments of the present invention in conjunction with Figs.3-6, it can be seen that the spatial channel coding architecture for practical systems can be designed according to the requirement for data transmission rate and the requirement for the number of transmit antennas and that of receive antennas. Table.1 lists the reachable data rate, spatial coding rate and diversity order for the spatial channel coding under different antenna configurations and modulations. The architectures labeled as SCC-I and SCC-II in Table.1, are respectively the architectures shown in Fig.4 and Fig.5 in the embodiments of the present invention. The spatial coding rate is the coding rate in the spatial dimension. For example, the case where four transmit antennas, two receive antennas and 16QAM mapping are used, is abbreviated as 4Tx-2Rx 16QAM. Assuming that the data rate required by the system is 4b/s/Hz, the data bits for one time coding once by the spatial channel encoder are 4 bits and 16-bit coded bits are obtained after being coded by each channel coding branch, then the spatial coding rate is 1/4. Next, after the 16 coding bits are processed through 16QAM mapping, the spatial channel coding architecture outputs 4 symbols corresponding to the 4 transmit antennas. For this case there are 2 receive antennas, so the diversity order is 8. From the table it can be seen that a rational spatial channel coding architecture in practical systems can be chosen according to the requirements for transmit antennas, data transmission rate and modulation mode, thus the UE can achieve high-speed data transmission under its limited receiver capability.

With regard to the spatial channel coding method proposed in the present invention, channel coding and space coding are combined to realize multi-channel parallel transmissions, so better system performance can be achieved while comparing with single-antenna transceiving architecture and existing BLAST techniques. This is verified in the simulation tests.

Table.1 The data rate, spatial coding rate and diversity order with different number of transmit and receive antennas

Table.2 illustrates the simulation parameters for different architectures with 8 different transmit and receive antennas. The 8 architectures are classified into 4 groups, as SISO (SISO: Signal Input Single Output) architecture, SCC architecture, BLAST architectures with and without (nBLAST) channel coding respectively Each group includes two specific architectures denoted as I indicating QPSK modulation and II indicating 8PSK modulation. In Table.2, SISO-I and SISO-II architectures don't perform interleaving, and adopt two coding modes respectively as (2, 1, 9) and (3, 1 , 9) to perform channel coding with coding rate as 1/2 and 1/3 respectively, and the two encoders both have 256 states. SCC-I and SCC-II listed in Table.2 are the parameters for the two embodiments in the present invention where the transmitter employs two or three transmit antennas respectively while the receiver has single receive antenna. With regard to the group of BLAST, the architectures for BLAST-I and BLAST-II are same as those for SISO-I and SISO-II.

The simulation tests for the above 8 architectures are based on the assumption that the propagation channel is quasi-static flat fading channel and the bit number for each frame is

130 during the data transmission, and other parameters in the simulation tests are set according to 3GPP UMTS FDD standards. The above 8 architectures are adopted for simulation tests to test the FER performance against the change of SNR. The simulation results are shown in Fig.7 and Fig.8.

Table.2 Simulation parameters

Fig.7 shows the curves for the FER of the received signals with change of the SNR corresponding to the four architectures as SISO-I, SCC-I, BLAST-I and nBLAST-I. Fig.8 shows the curves for the FER of the received signals with change of the SNR with regard to the four architectures as SISO-II, SCC-II, BLAST-II and nBLAST-II. In Fig.7 and Fig.8, the horizontal coordinate indicates the SNR of the receive signals and the vertical coordinate indicates the relevant FER. Referring to Fig.7, with the same FER, for example

FER=IO^"1, comparing the four system solutions, SNR of the received signals that can be tolerated by the system adopting SCC-I is the lowest and 4dB lower than that by SISO-I, 3dB lower than that by BLAST-I. Similarly, referring to Fig.8, with the same FER, for example FER=IO^"1, SNR of the received signals that can be tolerated by the system adopting SCC-II is also the lowest and IdB lower than that by SISO-II, 3dB lower than that by BLAST-II. Thus it can be concluded that SCC can achieve favorable system performance.

In simulation tests, also tested is the system data transmission rate, which is only an indication of the system spectrum efficiency, with unit as b/s/Hz, and the testing result is listed in the last row in Table.2. From the data transmission rate in Fig.2, it can be seen that the data transmission rate of the spatial channel coding scheme is much higher than that of the single antenna architecture. From Fig.7 and Fig.8, it is obvious that BLAST architecture has the poorest FER and has to use multiple receive antennas, although it provides the highest data transmission rate. Therefore, SCC is more suitable for providing downlink high-speed data service comparing with BLAST.

BENEFICIAL RESULTS OF THE INVENTION

Based on the above detailed descriptions of the embodiments of the present invention in conjunction with accompanying drawings, it can be seen that the spatial channel coding method, integrating channel coding and spatial coding as a whole solution, can implement multi-channel parallel transmissions when the number of receive antennas in the receiver is limited or even down to one, and thus increase the data transmission rate of the system, especially be very suitable for providing HPDS in downlink.

Moreover, the proposed spatial channel coding can have more transmit antennas and more channel coding states, for example 128 states, compared with existing STTC, to further improve the diversity gain and coding gain. Meanwhile, in designing the proposed spatial channel codes, the spatial channel codes can be determined flexibly according to the communication rate, communication quality, modulation mode and number of transmit antennas adopted by the transmitter, thus can be applied very easy and flexible. Additionally, the present invention also provide a judging criteria for the spatial channel codes generation matrix when the product of the number of transmit antennas and that of receive antennas is more than 3 is to maximize the minimum Euclidean distance, which is suitable for TCM coding and thus provides a feasible method for system extension in the future. While the present invention has been described with respect to embodiments thereof, those skilled in the art will recognize that the present invention is not limited to those specific embodiments described and illustrated herein. Various changes, modifications and equivalent arrangements may also be used to implement the invention without departing from the spirit and the scope of the invention. Accordingly, it is intended that the invention be limited only by the scope of the claims appended hereto.

Claims

CLAIMS:

1. A spatial channel coding method, comprising:

(a) inputting a group of serial data to be coded according to a predefined communication rate; (b) performing channel coding on said group of data with a corresponding coding criteria according to a predefined communication mode, to output multi-channel parallel coded signals, wherein there is relevant redundant information between each channel of coded signals;

(c) transmitting said multi-channel coded signals via a plurality of transmit antennas accordingly.

2. The spatial channel coding method according to claim 1, wherein step (b) includes: (bl) determining corresponding groups of branches for channel coding according to the number of said plurality of transmit antennas; (b2) determining branches for channel coding included in each group of branches according to a predefined modulation mode;

(b3) encoding said group of data inputted into said plurality of branches according to a relevant coding criteria.

3. The spatial channel coding method according to claim 2, wherein step (b2) further includes: determining a plurality of registers for channel coding in said branch according to requirement for communication quality; wherein said group of data inputted into said plurality of registers in said branch according to said relevant coding criteria are coded in step (b3).

4. The spatial channel coding method according to claim 3, wherein step (b3) includes: choosing data outputted from said corresponding plurality of registers in said plurality of branches according to said relevant coding criteria; generating coded data of said plurality of branches according to the outputted data.

5. The spatial channel coding method according to claim 4, wherein step (b3) further includes: performing symbol mapping on said plurality of branches of the generated coded data according to said modulation mode, to output said multi-channel parallel coded signals.

6. The spatial channel coding method according to claim 5, wherein said modulation mode at least is any one of BPSK, QPSK, 8PSK and QAM.

7. The spatial channel coding method according to claim 3, wherein if a product of the number of said transmit antennas and that of receive antennas is not more than 3 and the receive antennas at a receiving side are used to receive coded signals sent from said transmit antennas, said coding criteria makes that in a trellis diagram generated according to a codes generation matrix constructed by said plurality of branches and plurality of registers in each branch, a difference matrix constructed between output symbols corresponding to each decoding path can maximize the minimum of the rank/product of a matrix obtained through a product of the difference matrix and its conjugate transpose matrix.

8. The spatial channel coding method according to claim 7, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, said coding criteria makes that the minimum Euclidean distance in the trellis diagram is maximized when decoding is performed.

9. The spatial channel coding method according to claim 7, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, said coding criteria adopts conventional TCM (Trellis Code Modulation) coding.

10. The spatial channel coding method according to claim 1, wherein step (c) includes: multiplexing the coded signals spread using different spreading codes so as to transmit the multiplexed coded signals via the same antenna; transmitting each channel of multiplexed spread coded signals via said plurality of transmit antennas accordingly.

11. A spatial channel decoding method, comprising:

(a) receiving multi-channel parallel coded signals via at least one receive antenna, wherein said multi-channel parallel coded signals are transmitted via a plurality of corresponding transmit antennas at a transmitting side after spatial channel coding and there is relevant redundant information between each channel of coded signals;

(b) performing channel estimation on a plurality of wireless channels through which the coded signals are propagated according to received pilot signal;

(c) decoding the received signals by using the channel estimation result in accordance with spatial channel coding.

12. The spatial channel decoding method according to claim 11, wherein if the coded signals are received via a plurality of receive antennas in step (a), said plurality of wireless channels propagating the coded signals are estimated respectively according to the received pilot signal in step (b), and step (c) further includes: weighting said plurality of received coded signals by using the corresponding channel estimation result; decoding the weighted signals according to the spatial channel coding.

13. The spatial channel decoding method according to claim 11 or 12, wherein in step (c) a soft-decision Viterbi decoding method is used to decode the received signals.

14. The spatial channel decoding method according to claim 13, wherein said spatial channel coding is based on a communication mode adopted at the transmitting side.

15. The spatial channel decoding method according to claim 14, wherein said communication mode includes communication rate, communication quality, modulation mode and the number of transmit antennas.

16. The spatial channel decoding method according to claim 15, wherein if a product of the number of said transmit antennas and that of said receive antennas is not more than 3, a coding criteria adopted by said spatial channel coding makes that in a trellis diagram generated according to a codes generation matrix, a difference matrix constructed between output symbols corresponding to each decoding path can maximize the minimum of the rank/product of a matrix obtained through a product of the difference matrix and its conjugate transpose matrix.

17. The spatial channel decoding method according to claim 16, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, the coding criteria used by said spatial channel codes can maximize the minimum Euclidean distance in the trellis diagram when performing soft-decision Viterbi decoding.

18. The spatial channel decoding method according to claim 16, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, said spatial channel coding are TCM coding.

19. A spatial channel encoder, comprising: a plurality of encoding branches, for receiving a group of serial data to be coded respectively; wherein each encoding branch is composed of a plurality of registers, and performs channel coding on the group of received data by using a corresponding coding criteria according to a predefined communication mode; said plurality of encoding branches output parallel coded signals respectively and transmit them via a plurality of transmit antennas accordingly, and there is relevant redundant information between each channel of coded signals.

20. The spatial channel encoder according to claim 19, wherein said encoding branch is determined according to the number of said plurality of transmit antennas and a predefined modulation mode.

21. The spatial channel encoder according to claim 20, wherein the number of registers included in each encoding branch is based on requirement for communication quality, and coded data are generated from output data of the corresponding registers in each encoding branch according to the relevant coding criteria.

22. The spatial channel encoder according to claim 21, wherein said modulation mode at least is any one of BPSK, QPSK, 8PSK and QAM.

23. The spatial channel encoder according to claim 19, wherein if a product of the number of said transmit antennas and that of receive antennas is not more than 3 and the receive antennas are at a receiving side for receiving coded signals sent from said transmit antennas, said coding criteria makes that in a trellis diagram generated according to a codes generation matrix constructed by said plurality of branches and a plurality of registers in each branch, a difference matrix constructed between output symbols corresponding to each decoding path can maximize the minimum of the rank/product of a matrix obtained through a product of the difference matrix and its conjugate transpose matrix.

24. The spatial channel encoder according to claim 19, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, said coding criteria makes that the minimum Euclidean distance in the trellis diagram can be maximized when decoding is performed.

25. The spatial channel encoder according to claim 19, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, said coding criteria adopts conventional TCM (Trellis Code Modulation) coding.

26. The spatial channel encoder according to claim 19, further comprising: a plurality of multiplexing units, each of which is used for multiplexing the coded signals spread with different spreading codes, so as to transmit the multiplexed coded signals via the same antenna; wherein said multiple transmit antennas transmit each channel of multiplexed spread coded signals accordingly.

27. A spatial channel decoder, comprising: a decoding module, for decoding received signals by using channel estimation result in accordance with spatial channel coding; wherein the received signals are multi-channel parallel coded signals received via at least one receive antenna, said multi-channel parallel coded signals are transmitted via a plurality of corresponding transmit antennas at a transmitting side after spatial channel coding and there is relevant redundant information between each channel of coded signal; the channel estimation result is obtained by at least one channel estimation unit through performing channel estimation on a plurality of wireless channels propagating the coded signals in accordance with received pilot signal.

28. The spatial channel decoder according to claim 27, further comprising: a weighting module, for weighting a plurality of received signals using the corresponding channel estimation result obtained by a plurality of channel estimation units according to the pilot signal received by a plurality of receive antennas; wherein the decoding module performs soft-decision Viterbi decoding on the weighted signals in accordance with spatial channel coding.

29. The spatial channel decoder according to claim 28, wherein said spatial channel coding is based on a communication mode adopted at the transmitting side, the communication mode comprising communication rate, communication quality, modulation mode and the number of transmit antennas.

30. The spatial channel decoder according to claim 29, wherein if a product of the number of said transmit antennas and that of said receive antennas is not more than 3, a coding criteria adopted by said spatial channel coding makes that in a trellis diagram generated according to a codes generation matrix, a difference matrix constructed between output symbols corresponding to each decoding path can maximize the minimum of the rank/product of a matrix obtained through a product of the difference matrix and its conjugate transpose matrix.

31. The spatial channel decoder according to claim 30, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, the coding criteria adopted by said spatial channel coding is to maximize the minimum

Euclidean distance in the trellis diagram when performing soft-decision Viterbi decoding.

32. The spatial channel decoder according to claim 30, wherein if the product of the number of said transmit antennas and that of said receive antennas is more than 3, said spatial channel coding is TCM (Trellis Code Modulation) coding.

33. A wireless network system, comprising: a spatial channel encoder, for performing channel coding on serial data to be transmitted, to output multi-channel parallel coded signals, wherein there is relevant redundant information among each channel of coded signals; a plurality of transmit antennas, for transmitting each channel of coded signals accordingly.

34. The wireless network system according to claim 33, wherein the spatial channel encoder comprises: a plurality of encoding branches, for receiving a group of serial data to be coded respectively; wherein each encoding branch is composed of a plurality of registers and performs channel coding on the group of received data according to a predefined communication mode by using corresponding coding criteria; said plurality of encoding branches output parallel coded signals respectively to transmit them via said plurality of transmit antennas accordingly, and there is relevant redundant information between each channel of coded signals.

35. The wireless network system according to claim 34, wherein said encoding branches are determined by the number of said plurality of transmit antennas and predefined modulation mode, the number of registers included in each encoding branch is determined by requirement for communication quality, and said coded data are generated from output data of the corresponding registers in each encoding branch according to the relevant coding criteria.

36. A wireless terminal, comprising: at least one receive antenna, for receiving multi-channel parallel coded signals, wherein said multi-channel parallel coded signals are transmitted via a plurality of transmit antennas at a transmitting side after spatial channel coding, and there is relevant redundant information between each channel of coded signals; at least one channel estimation unit, for performing channel estimation on a plurality of wireless channels propagating the coded signals according to received pilot signal; a spatial channel decoder, for decoding the received signals by using the channel estimation result according to the spatial channel coding.

37. The wireless terminal according to claim 36, further comprising: a weighting module, for weighting said plurality of received coded signals using the corresponding channel estimation result obtained by said plurality of channel estimation units according to the pilot signal received by said plurality of receive antennas; wherein the spatial channel decoder performs soft-decision Viterbi decoding on the weighted signals according to the spatial channel coding.

38. The wireless terminal according to claim 37, wherein said spatial channel coding is based on a communication mode adopted at the transmitting side, the communication mode comprising communication rate, communication quality, modulation mode and the number of transmit antennas.