Data transmission apparatus for DS/CDMA system equipping MIMO antenna system

[0001] The present invention relates to a Direct Sequence DS/Code Division Multiple Access (CDMA) communication system equipping a multiple-input multiple-output (MIMO) antenna; and, more particularly, to a data transmission apparatus for detecting a transmission data by linear operation of receiving data with a diversity gain of a theoretical maximum order of receiving diversity for the CDMA system equipping MIMO antenna in frequency selective fading channels.

[0002] Conventional data transmission apparatus and methods for simultaneously transmitting multiple data by using a plurality of transmission antennas are introduced by G.J Foschini et al. in Simplified processing for high spectral efficiency wireless communication employing multi-element arrays, IEEE J. Selet. Areas Commun., vol. 17, pp. 1841-1852, Nov. 1999 (Hereinafter Foschini).

[0003] The Foschini teaches a V-BLAST method as a receiving method for receiving transmitting data from a plurality of transmitting antennas by increasing an order of diversity in order from data having bigger signal-to-noise ratio with a plurality of receiving antennas in a time division multiple access (TDMA) or frequency division multiple access (FDMA) system with frequency flat fading environment. In this method, the number of receiver antennas should be larger than the number of transmission antennas. The V-BLAST receiving method uses a concept that channel weights between the transmitting antennas and the receiving antennas are independent. Generally, the receiver detects data based on the number of receiving antennas (1 - the number of transmitting antennas)/2 orders of receiving diversity.

[0004] In a mean time, general adaptive interference control methods for receiving antenna system are disclosed by S.C. Hong, J.H. Choi, S.R. Kim and Y,H. Lee, in Constrained MMSE-RAKE receivers for DS/CDMA systems in frequency selective fading channels, in Proc. VTC02 spring, May 2002.

[0005] For using a concept of the V-BLAST receiving method in a code division multiple access (CDMA) system, H. Huang et al. introduces another transmission apparatus and method in Achieving high data rates on the UMTS downlink shared channel using multiple antennas, in Proc. International Conference on 3G Mobile Communication Technologies, Mar. 2001(Hereinafter Huang). The Huang's method are considered as stand transmission method for high speed data transmission in 3rd generation mobile communication standard at 3GPP TR 25.848 (V 4.0.0), Physical layer aspects of UTRA High Speed Downlink Packet Access (Release 4), Mar. 2001 (Hereinafter 3GPP). The Huang's transmitting method simultaneously transmits multiple data sequences by spreading the data sequences, which are orthogonal to each others. After spreading the data sequences, spread data sequences are scrambled by using one scramble code for maintaining the orthogonality between spread data.

[0006] The transmitting and receiving method introduced by Huang and Foschini is designed and developed for frequency flat fading environment and cannot obtain full order of diversity. For frequency selective fading channels, A. Lozano and C. Papadias introduces a V-BLAST receiver combined with decision feedback equalizer (DFE) in Space-time receiver for wideband BLAST in rich scattering wireless channels, in Proc. IEEE VTC00, Japan, May 2000 (Hereinafter Lozano) and similar method of Lozano is also announced by Xu Zhu and R.D. Murch in MIMO-DFE based BLAST over frequency selective fading channels, in Proc. GLOBECOM0l, Nov. 2001(Hereinafter XU). The method of Lazano and XU implement the V-BLAST receiving method of Foschini by using the DFE for gathering multipath energy in a time division multiple access system.

[0007] In the code division multiple access (CDMA) system, the decision feedback equalizer may be impossible to use because of a chip unit signal which has comparatively low signal-to-noise ratio (SNR) and high implementation complexity. Therefore, there is no a receiving method has been introduced for the DS/CDMA system equipping MIMO antenna in frequency selective fading channel environment.

[0008] It is, therefore, an object of the present invention to provide a transmission system for a direct sequence/code division multiple access (DS/CDMA) system using a multiple-input multiple-output (MIMO) antenna.

[0009] It is another object of the present invention to provide a transmitting system in frequency selective fading environment for the DS/CDMA system equipping the MIMO antenna by applying different scramble code to each transmitting data according to corresponding transmitting antenna.

[0010] It is still another object of the present invention to provide a receiving system in frequency selective fading environment for the DS/CDMA system equipping the MIMO antenna by obtaining diversity gain of a maximum order of diversity in order to effectively detect data.

[0011] In accordance with an aspect of the present invention, there is provided a transmitting apparatus for a direct sequence (DS) / code division multiple access (CDMA) equipping multiple-input multiple-output antenna system, including: serial to parallel converting unit for converting serially inputted high speed data to parallel data sequence; spreading unit for generating spread data by spreading the parallel data sequence based on a spreading code, which is orthogonal to the parallel data sequence; pilot signal generating unit for generating a pilot signal; combining unit for generating combined data by combining the spread data and the pilot signal; scrambling unit for scrambling the combined data by using different scrambling code in order to generate scrambled data; and a plurality of transmitting antennas for independently transmitting the scrambled data.

[0012] In accordance with an aspect of the present invention, there is also provided a receiving apparatus for a direct sequence (DS) / code division multiple access (CDMA) equipping multiple-input multiple-output antenna system, including; a plurality of receiving antennas for receiving a receiving analog signal transmitted from a transmitter; matching filtering unit for converting the receiving analog signal to a digital signal; input signal generating unit for gathering transmitting data in the digital signal and outputting the gathered transmitting data; a plurality of de-spreading unit of which the number of the de-spreading unit are in linearly proportion to the number of paths in multipath for gathering energy, the number of transmitting antennas in the transmitter and the number of the spreading codes for computing an scalar product value of a spreading code and a scramble code in the transmitting data; channel estimating unit including a plurality of channel estimators as many as the number of the receiving antennas for estimating channels based on a pilot signal from the de-spreading unit; combining unit for gathering energies of transmitting data by using an output of de-spreading unit and the channel estimating unit; and data deciding unit for deciding an estimation value of transmitting symbol by using the combined signal from the combining unit.

[0013] In accordance with another aspect of the present invention, there is also provided a receiving apparatus for a direct sequence (DS) / code division multiple access (CDMA) equipping multiple-input multiple-output antenna system, including; a plurality of receiving antennas for receiving a receiving signal transmitted from a transmitter; input signal generating unit for gathering transmitting data in the receiving signal and outputting the gathered transmitting data; adaptive filtering unit for filtering a complex receiving signal according to a tab weight, which is controlled in a predetermined cycle and outputting a filtered data; channel estimating unit for estimating a phase and an amplitude of a channel between one of transmitting antennas and one of receiving antennas based on the filtered data and outputting an estimation data as a result of estimation; signal recovering unit for recovering the receiving signal to an original signal by combining the estimation data and filtered data with all components of multipath; selecting unit for selecting one of the original signal and a training data signal supplied by the receiver and outputting a selected signal; reference signal generating unit for generating a reference signal based on the selected signal and the estimation data; error calculating unit for calculating error between the filtered data and the reference signal by comparing the filtered data and the reference signal; tab weight controlling unit for controlling the tab weight of the adaptive filtering unit based on a constraint MMSE criterion having at least one condition.

[0014] In accordance with still anther aspect of the present invention, there is also provided a transmission apparatus for a direct sequence (DS) / code division multiple access (CDMA) equipping multiple-input multiple-output antenna system, including; serial to parallel converting unit for converting serially inputted high speed data to parallel data sequence; spreading unit for generating spread data by spreading the parallel data sequence based on a spreading code, which is orthogonal to the parallel data sequence; pilot signal generating unit for generating a pilot signal; combining unit for generating combined data by combining the spread data and the pilot signal; scrambling unit for scrambling the combined data by using different scrambling code in order to generate scrambled data; a plurality of transmitting antennas for independently transmitting the scrambled data; a plurality of receiving antennas for receiving the scrambled data, which is a analog signal transmitted from a plurality of the transmitting antennas; matching filtering unit for converting the analog signal to a digital signal; input signal generating unit for gathering transmitting data in the digital signal and outputting the gathered transmitting data; a plurality of de-spreading unit of which the number of the de-spreading unit are in linearly proportion to the number of paths in multipath for gathering energy, the number of transmitting antennas in the transmitter and the number of the spreading codes for computing an scalar product value of a spreading code and a scramble code in the transmitting data; channel estimating unit including a plurality of channel estimators as many as the number of the receiving antennas for estimating channels based on a pilot signal from the de-spreading unit; combining unit for gathering energies of transmitting data by using an output of de-spreading unit and the channel estimating unit; and data deciding unit for deciding an estimation value of transmitting symbol by using the combined signal from the combining unit.

[0015] In accordance with further still aspect of the present invention, there is also provided 8. A transmission apparatus for a direct sequence (DS) / code division multiple access (CDMA) equipping multiple-input multiple-output antenna system, including; serial to parallel converting unit for converting serially inputted high speed data to parallel data sequence; spreading unit for generating spread data by spreading the parallel data sequence based on a spreading code, which is orthogonal to the parallel data sequence; pilot signal generating unit for generating a pilot signal; combining unit for generating combined data by combining the spread data and the pilot signal; scrambling unit for scrambling the combined data by using different scrambling code in order to generate scrambled data; a plurality of transmitting antennas for independently transmitting the scrambled data; a plurality of receiving antennas for receiving the scrambled data transmitted from a transmitting antennas; input signal generating unit for gathering transmitting data in the receiving signal and outputting the gathered transmitting data; adaptive filtering unit for filtering a complex receiving signal according to a tab weight, which is controlled in a predetermined cycle and outputting a filtered data; channel estimating unit for estimating a phase and an amplitude of a channel between one of transmitting antennas and one of receiving antennas based on the filtered data and outputting an estimation data as a result of estimation; signal recovering unit for recovering the receiving signal to an original signal by combining the estimation data and filtered data with all components of multipath; selecting unit for selecting one of the original signal and a training data signal supplied by the receiver and outputting a selected signal; reference signal generating unit for generating a reference signal based on the selected signal and the estimation data; error calculating unit for calculating error between the filtered data and the reference signal by comparing the filtered data and the reference signal; tab weight controlling unit for controlling the tab weight of the adaptive filtering unit based on a constraint MMSE criterion having at least one condition.

[0016] The above and other objects and features of the present invention will become apparent from the following description of the preferred embodiments given in conjunction with the accompanying drawings, in which:

[0017] FIG. 1 is a block diagram illustrating a transmission system for a code division multiple access (CDMA) system equipping a multiple-input multiple-output (MIMO) antenna in accordance with a preferred embodiment of the present invention;

[0018] FIG. 2 is a diagram showing a receiver in accordance with the present invention. The receiver in FIG. 2 implements a conventional rake receiver by enhancing some parts of it according to the present invention;

[0019] FIG. 3 is a diagram depicting a receiver in accordance with another embodiment of the present invention;

[0020] FIG. 4 is a diagram showing a receiver in accordance with another preferred embodiment of the present invention;

[0021] FIG. 5 is a detailed diagram illustrating an adaptive filter block 36 and 46 in FIGS. 3 and 4 in accordance with the present invention;

[0022] FIG. 6 is a diagram depicting an adaptive filter block in FIGS. 3 and 4 in accordance with another embodiment of the present invention;

[0023] FIG. 7 is a detailed diagram showing an adaptive filter bank block in accordance with a preferred embodiment of the present invention;

[0024] FIG. 8 is a detailed diagram showing a data decider in accordance with the present invention;

[0025] FIG. 9 is a graph showing a bit error rate of the present invention with 4 transmitting antennas and 4 receiving antennas in frequency flat fading channel environment;

[0026] FIG. 10 is a graph depicting a receiving diversity gain obtained in the present invention; and

[0027] FIG. 11 is a graph showing a performance of the present invention in three frequency selective fading channel environment.

[0028] Other objects and aspects of the invention will become apparent from the following description of the embodiments with reference to the accompanying drawings, which is set forth hereinafter.

[0029] FIG. 1 is a block diagram illustrating a transmission system for a code division multiple access (CDMA) system equipping a multiple-input multiple-output (MIMO) antenna in accordance with a preferred embodiment of the present invention.

[0030] Referring to FIG. 1, the transmission system includes a transmitter and a receiver. The transmitter has a serial/parallel converter 11, a plurality of spreaders 12, a combiner 13, a scrambling processor 14 and a plurality of transmission antenna 15. The receiver contains a plurality of receiving antennas 16 and a receiver 17.

[0031] The MIMO antenna can simultaneously transmit data as much as multiple times of data transmitted by one antenna at once. Therefore, high speed transmission data is used as an input data of the transmitting system for explaining the present invention in below.

[0032] The serial/parallel converter 11 receives high speed transmission data and converts the high speed transmission data to K parallel data sequences. Each of K parallel data sequences includes P transmission data, wherein K is the number of spreading codes used for transmitting and P is the number of transmission antennas. The KXP transmission data are divided to P data symbols of K groups and passed to a plurality of spreader 12.

[0033] Each of a plurality of spreaders 12 receives corresponding parallel data sequence and spreads the parallel data sequence based on one of corresponding spreading code C_k to spreading code C_k, which are orthogonal to corresponding parallel data sequence. The spread data of each spreader 12 is selectively passed to one of corresponding combiners 13. In a meantime, 0^th spreader 12 receives a pilot symbol and 0^th spreading code. The pilot symbol is spread based on the 0^th spreading code.

[0034] Each of a plurality of the combiner 13 receives the spread data from multiple of corresponding spreaders 12 and combines the spread data and pilot symbol, wherein the pilot symbol is combined for estimating a channel according to each of the transmission antennas.

[0035] A plurality of the scrambling processors 14 receives the combined code from the combiners 13. Each of the scrambling processors 14 scrambles the combined data by using different scramble code according to corresponding transmitting antenna. The scrambled data is passed to corresponding antenna and transmitted to the receiver.

[0036] As mentioned above, using different scrambling code according to the transmitting antenna is distinguishable to conventional transmitter.

[0037] In the present invention, different scrambling codes are used for scrambling the combined data according to each of transmission antennas. In the conventional method for transmitting data in the MIMO antenna system, one scrambling code is used for all transmission antennas.

[0038] For explaining difference of the conventional transmitting method using one scrambling code and the present invention using different scrambling codes, it is necessary to define a scrambling operator.

[0039] If there are two matrices A and B having same dimension, the scrambling operator produce a result matrix by multiplying each of entries of A and corresponding entries of B. For example, in case of two N1 matrices such as a matrix Aa₁, .., a_n^T and a matrix Bb₁, ..., b_n^T, the result matrix of A B is a₁b₁, a₂b₂..., a_nb_n^T. That is, A B a₁b₁, a₂b₂, ..., a_nb_n ^T.

[0040] The transmitting method of the present invention and conventional method are compared and explained by using an exemplary apparatus having two transmission antennas, two spreading codes C₁ and C₂ and four data d₁ to d₄.

[0041] The conventional method produces transmitting data y defined by following equation in case of frequency flat fading channel.

[0042] ,wherein h₁ and h₂ are channel parameters.

[0043] Base on d₁, a de-spread signal in a receiver side will be expressed as following equation.

[0044] Referring to Eqs. 1 and 2, there is a strong interference signal d₂ left on data symbol after transmitting the scrambled data by one scrambling code. As shown in Eq. 2, the channel weight of each signal is different between the scrambled signal d₁ and the interference signal d₂. Therefore, based on difference of the channel weight, the interference signal is can be eliminated by using a decorrelating receiving method, a minimum mean square error method or V-BLAST receiving method, which are all conventional receiving methods.

[0045] In the above mentioned case, the number of receiving antennas must be larger than the number of transmitting antennas, and the channel parameters have to be independent. Furthermore, full order of receiving diversity can not be obtained by the above conventional receiving methods generally used in the MIMO antenna system. Among them V-BLAST method gives the a maximum order of receiving diversity in conventional method is M-(P-l)/2, wherein M is the number of receiving antennas and P is the number of transmitting antennas. Since the theoretical maximum order of diversity is M, there is a loss of order of diversity as much as (P-l)/2.

[0046] The interference from the data spread by different spreading codes can be eliminated by using orthogonality between codes in case of frequency flat fading channel, however, in case of frequency selective fading channels, the orthogonality between different spreading codes is not maintained. Therefore, the interference in the frequency selective fading channel environment cannot be eliminated.

[0047] For eliminating the interference in the frequency selective fading channel environment, use of decision feedback equalizer can be considered. However, the decision feedback equalizer can not work in the code division multiple access (CDMA) environment, because the signal-to-noise ratio of a chip unit signal is less than 0 decibel (dB). Therefore, there is no conventional apparatus for MIMO DS/CDMA systems in the frequency selective fading channel environments.

[0048] In contrast, the scrambled data, which is transmitting data, of the present invention can be defined as following equation.

[0049] Based on d₁, a de-spread signal in a receiver can be expressed as following equation.

[0050] Referring to Eq. 4, d₂ is spread based on different spreading code and transmitted from same antenna comparing to other signals such as di, d₃, d₄. d₄ is spread by an orthogonal spreading code and transmitted from different antenna. The interference of d₂ is completely eliminated and the interference of d₄ is added. However, the size of the interference from d₄ is quite small because a correlation value between different scrambling codes is very small.

[0051] As shown in Eqs. 3 and 4, all data simultaneously transmitted has different signature sequence and the data in Eq. 4 are very similar to the signals in CDMA with one transmission antenna system. Therefore, various interference elimination methods for the CDMA system with one transmission antenna like adaptive interference suppression method disclosed by S. C. Hong, J. H.Choi, S. R. Kim and Y. H.Lee, in Constrained MMSE-RAKE receivers for DS/CDMA systems in multipath fading channels, in Proc. VTC02 spreing, May 2002, can be applied to detect the transmitted data from the transmitter of current invention.

[0052] Also, the present invention has excellent detection performance with simple linear calculation. In the present invention, since power of interference signal is very small due to small correlation value between codes, a conventional rake receiver which can obtain complete order of diversity gain, but does not eliminates interferences shows good detection performance.

[0053] Hereinafter, difference of the conventional method using one scrambling code and the present invention using multiple different scrambling codes in frequency selective fading channel environment is explained as followings.

[0054] A fading channel of multipath at time t can be defined as following equation. $\begin{array}{cc}{h}_{\left(t\right)}=\underset{P=1}{\overset{P}{}}\underset{m=1}{\overset{M}{}}\underset{l=1}{\overset{L}{}}{h}_{\left(t\right),({\mathrm{pml}}^{}\left(l-{}_1\right)}& \mathrm{Eq}.5\end{array}$

[0055] In Eq. 5, h_(t),pml⁶1₁) is a channel weight of 1^th multipath from p^th transmission antenna to m^th receiving antenna. ₁ is a transmission delay of 1^th multipath and for the convenience ₁ is assumed as (l1)^T_c, wherein T_c is a cycle of a chip.

[0056] If it assumes that a channel is not changed within a period of transmitting each code block of parallel data sequence, the channel weight is maintained as h_(l),pmlh_(l),pml(n) within nT<t<(n1)T , when a symbol cycle is T. Nx1 k^th spreading code vector is c_kc_kl,... c_kN^t and Nl p^th normalized scrambling code vector is s_pS_pl,...,S_pN^t. 0^th spreading code c₀ is allocated to a pilot symbol and is used for spreading the pilot symbol d_0p(n) according to p. A spreading code of 1^th path is defined as c_kl0^t_l1c^t_k,0^t_l1^t and normalized scrambling code of 1^th path is defined as s_pl0^t_l1s^t_p0^t_l1^t. A vector r_m(n) of nth transmitting data sequence among outputs of a chip matching filter according to a receiving signal of m^th receiving data is defined as following equation. $\begin{array}{cc}\begin{array}{c}{r}_m\left(n\right)=\underset{k=1}{\overset{K}{}}\underset{P=1}{\overset{P}{}}\underset{l=1}{\overset{L}{}}{c}_{\mathrm{kl}}{s}_{\mathrm{pl}}{h}_{\mathrm{pm},l}\left(n\right)d_{\mathrm{kp}}\left(n\right)+u_m\left(n\right)\\ =\underset{k=1}{\overset{K}{}}\underset{p=1}{\overset{P}{}}\underset{l=1}{\overset{L}{}}{X}_{\mathrm{kpl}}{h}_{\mathrm{pm},l}\left(n\right)d_{\mathrm{kp}}\left(n\right)+u_m\left(n\right)\end{array}& \mathrm{Eq}.6\end{array}$

[0057] d_kp(n) is data symbol, which is transmitted from pth antenna after spreading with k^th spreading code, u_m(n) is a sum of background noise and all interference signals including interference signal of other user and neighbor symbol of m^th antenna. For convenience, C_kls_pl is defined as X_kpl.

[0058] As mentioned above, the transmitter of the present invention simultaneously transmits a plurality of scrambled data with using multiple different scramble code to a receiving. Advantages of applying multiple different scramble code are explained in later with a receiver of the present invention by referring to FIG. 2.

[0059] FIG. 2 is a diagram showing a receiver in accordance with the present invention. The receiver in FIG. 2 implements a conventional rake receiver by enhancing some parts of it according to the present invention.

[0060] Referring to FIG. 2, the receiver receives signals from a transmitter in FIG. 1 in accordance with the present invention when there are L multipath and M receiving antennas.

[0061] Referring to FIG. 2, the rake receiver includes: M chip matching filters 22; an input buffer 23; (K1)PLM de-spreaders 24 for gathering multipath energies of each transmitting data; M channel estimators 25 for estimating a channel based on output of de-spreaders 23 according to a pilot signal at the receiving antenna 21; a combiner 26 for combining energies of each transmitting data based on a result of channel estimation and the output of de-spreaders and a data decider 27.

[0062] Each of de-spreaders 24 performs calculation of a scalar product value of r_m(n) , a spreading code and a scrambling code in corresponding components of multipath of each transmitting data.

[0063] For example, if the de-spreader 24 for de-spreading first components of multipath of d₁₁(n), which is transmitted through a first antenna and spread by a first spreading code, calculates X_11,1^Hr_m(n) and pass it to the combiner 26.

[0064] The channel estimator 25 estimates PL channel parameters by using outputs of de-spreader 24.

[0065] The combiner 26 combines receiving energies of KP data symbols d₁₁(n)d_kp(n) , which are simultaneously transmitted at the transmitter.

[0066] The combiner 26 generates a combined energy data as following equation with d_kp(n) as input and outputs to the data decider 27. $\begin{array}{cc}{\stackrel{\_}{d}}_{\mathrm{kp}}\left(n\right)=\underset{m=1}{\overset{M}{}}\underset{l=1}{\overset{L}{}}{\hat{h}}_{\mathrm{pm},l}^{*}{X}_{\mathrm{kp},l}^H{r}_m\left(n\right)& \mathrm{Eq}.7\end{array}$

[0067] The data decider 27 decides data symbol closest to a constellation by using the combined energy data of Eq. 7 and generates final estimation value of transmitted data symbol. It is explained in detail in later with FIG. 8.

[0068] As mentioned above, since the transmitter in FIG. 1 transmits data by using the multiple different scramble codes, the receiver in FIG. 2 receives the data having very similar characteristics to data received in a conventional receiver in frequency flat fading channel environment, therefore, various kind of conventional receiver can be implemented for the present invention.

[0069] However, in case of using the receiver in FIG. 2, which de-spreads receiving signal by using a spreading code, not only additive white Gaussian noise (AWGN) but also interference between other data may be added into the de-spread data. As a result, a receiving performance may be decreased.

[0070] For overcoming degrading of the receiving performance, an adaptive interference control method can be used. The adaptive interference control method de-spreads receiving signal by using optimum adaptive filter weight based on a constrained minimum mean square error instead of using the spreading code.

[0071] FIGS. 3 and 4 show receivers in accordance with another embodiment of the present invention. The receivers in FIGS. 3 and 4 use the adaptive interference control method for increasing a receiving performance by effectively eliminating interference signals

[0072] The receivers includes an input signal generating unit, an adaptive filtering unit, a channel estimating unit, a signal recovering unit, a selecting unit, a reference signal generating unit, an error calculating unit and a tab weight controlling unit.

[0073] The input signal generating unit gathers transmitting data in a receiving signal and passes the gathered transmitting data to the adaptive filtering unit.

[0074] The adaptive filtering unit filters a complex receiving signal according to a tab weight, which is controlled in a predetermined cycle and passes a filtered data to the channel estimating unit.

[0075] The channel estimating unit estimates a phase and an amplitude of a channel between one of transmitting antennas and one of receiving antennas based on the filtered data and output an estimation data as a result of estimation.

[0076] The signal recovering unit recovers the receiving signal to an original signal by combining the estimation data and filtered data with all components of multipath.

[0077] The selecting unit selects one of the original signal and a training data signal supplied by the receiver and output a selected signal.

[0078] The reference signal generating unit generates a reference signal based on the selected signal and the estimation data.

[0079] The error calculating unit calculating error between the filtered data and the reference signal by comparing the filtered data and the reference signal.

[0080] The tab weight controlling unit controls the tab weight of the adaptive filtering unit based on a constraint MMSE criterion having at least one condition.

[0081] The receivers in FIGS. 3 and 4 are identical except a channel estimation method and adaptive filter weight update algorithm. Therefore, each of receivers are explained in focus on the difference of two receivers.

[0082] FIG. 3 is a diagram depicting a receiver in accordance with another embodiment of the present invention.

[0083] Referring to FIG. 3, the receiver includes a receiving antenna 31, a chip matching filter 32, a input buffer 33, a error generation block 34, a simple mean channel estimator 35, an adaptive filter block 36, a combiner 37, a data decider 38 and a reference signal selector 39.

[0084] The receiving antenna 31 receives analog signals from the transmitter and the chip matching filter 32 converts the analog signals to digital signals.

[0085] The input signal generating buffer 33 outputs vector r_m(n) as a gathered transmitting data to corresponding adaptive filter block 36 by gathering n^th transmitting data sequence among digital signals from the chip matching filter 32.

[0086] The each of adaptive filtering block 36 includes a multi-condition filter weight updating block 51 and a plurality of adaptive filter bank 52 in FIG. 5, or includes one condition filter weight updating block 61 and a plurality of adaptive filter bank block 62.

[0087] In the receiver in FIG. 3, a standard adaptive filter weight is w_kpml,l(n) for detecting a 1^st multipath component of data symbol at m^th receiving antenna, wherein the data symbol is transmitted by being spread based on k^th spreading code from p^th transmitting antenna, a proposal object function is applies. The proposal object function is defined as following equation.

[0088] wherein s_pS_Pl,...,S_pl , g_l are L1 vector having 1 as 1^th element and 0 as all others.

[0089] The condition applied to the Eq. 8 is a condition for making an inter-path interference (IPI) as 0 by maintaining a scalar product value of a spreading code of 1^th multipath and a scalar product value of the adaptive filter weight and spreading code of each multipath as 1 and maintaining an scalar product value of other multipath components as 0.

[0090] By eliminating the interference between multipath based on the condition, the receiving performance is increased since channel is estimated without bias although the simple mean channel estimation method is used based on the adaptive filter's output.

[0091] By using Langlangian multiplier, w_kpm,l(n) can be calculated from Eq. 8, wherein w_kpm,l(n) minimizes the object function Eq. 8. It is expressed as following equation.

[0092] In Eq. 9, R_mEr_m(n)r^H_m(n) . Therefore, Eq. 9's w⁰_kpml can be used as w_kpml(n) by eliminating k and a speed of convergence of adaptive filter is increased by updating adaptive filter weight by combining error signals of all k.

[0093] The error generation block 34 calculates errors between a reference signal and output of corresponding adaptive filter, which is a filtered data. The calculated errors are passed to the filter weight updating block 51 or 61 of the adaptive filter block 36. An error signal vector of 1^th path of m^th receiving antenna is e_ml(n)e^T_1ml,le^T_2ml,...,e^T_pm,l^T and the error signal vector of data of each transmitting antennas is e_ml(n)e_0pm,l,e_1pm,l,...,e_kpm,l^T, wherein p, m, l, k is expressed as following equation.

[0094] In Eq. 10, h_pm,l(n) is an estimation value of 1^th multipath channel between p^th transmitting antenna and m^th receiving antenna. It is passed from the simple mean channel estimation block 35 in FIG. 3.

[0095] The adaptive filter block 36 updates adaptive filter weight by using following equation according to each transmitting antenna index p(1<p<P). $\begin{array}{cc}{w}_{\mathrm{pm},l}\left(n+1\right)=w_{\mathrm{pm},l}\left(n\right)+P_{s_p}^{}\left(\underset{k=0}{\overset{K}{}}{e}_{\mathrm{kpm},l}^{*}\left(n\right)r_m\left(n\right)\right)& \mathrm{Eq}.11\end{array}$

[0096] In Eq. 11, P_sp is defined as Is_p(s^H_ps_p)¹s^H_p and it is orthogonal projection vector, which is projected to a space orthogonal to s_p.

[0097] The adaptive filter bank 52 or 62 in the adaptive filter block 36 scrambles the spreading code of symbol corresponding to simple adaptive filter weight as an adaptive filter weight for a spread data. As a result, y_pm,l(n)y_0pm,ly_1pm,l...,y_kpm,l^T is outputted. A vector gathering outputs of the adaptive filter bank for each transmitting antenna is defined as y_m,l(n)y^T_1pmly^T_2pml ...,y^T_pm,l^T.

[0098] The receiver in FIG. 3 estimates a channel by using below equation which represents a simple mean channel estimation method. $\begin{array}{cc}{\hat{h}}_{\mathrm{pm},l}\left(n\right)=\frac{1}{2N_p}\underset{i=1}{\overset{N_p}{}}{d}_{0p}^{*}\left(n-i\right)w_{\mathrm{pm},l}^H\left(n-i\right)r_m\left(n-i\right)& \mathrm{Eq}.12\end{array}$

[0099] Referring to Eq. 12, the final channel estimation value is calculated by computing a mean of temporally channel estimation value based on a plurality of pilot symbol for reducing affection of noise. N_p is the number of pilot signal used for estimating the channel.

[0100] The final channel estimation value calculated by Eq. 12 is unbiased and it is a maximum likelihood channel estimation value under a condition of estimating the channel by using an adaptive filter output, which has a plurality of conditions. In the Eq. 12, d_0p(n) is a pilot symbol transmitted from p^th antennal.

[0101] The combiner 37 generates decision variable for each of KP data symbols d₁₁(n)d_Kp(n) . For d_KP(n) , the combiner 37 generates a combined signal and the combined signal is outputted to the data decider 38. The combined signal is expressed as following equation. $\begin{array}{cc}{\stackrel{\_}{d}}_{\mathrm{kp}}\left(n\right)=\underset{m=1}{\overset{M}{}}\underset{l=1}{\overset{L}{}}{{\hat{h}}_{\mathrm{pm},l}^{*}\left(c_{\mathrm{ki}}{w}_{\mathrm{pm},l}\right)}^H{r}_m\left(n\right)& \mathrm{Eq}.13\end{array}$

[0102] The data decider 38 decides data symbol closest to a constellation by using the combined signal and output a final estimation value of data symbol.

[0103] The reference signal selector 39 receives the final estimation value and generates a pilot combined signal by combining the final estimation value with pilot symbol. The pilot combined signal is passed to the channel estimator 35 and error generation block 34.

[0104] FIG. 4 is a diagram showing a receiver in accordance with another preferred embodiment of the present invention.

[0105] Referring to FIG. 4, the receiver includes a plurality of receiving antennas 41, a chip matching filter 42, an input buffer 43, an error generation block 44, a maximum likelihood channel estimator 45, an adaptive filter block 46, a combiner 47, a data decider 38 and reference signal selector 39.

[0106] As mentioned above, the receiver in FIG. 4 is very similar to the receiver in FIG. 3, therefore, overlapped elements are not explained and omitted.

[0107] In the receiver in FIG. 4, a standard adaptive filter weight is w_kpm,l(n) for detecting a 1^th multipath components of data symbol at m^th receiving antenna, wherein the data symbol is transmitted by being spread based on k^th spreading code from p^th transmitting antenna, a proposal object function is applies. The proposal object function is defined as following equation.

[0108] A condition applied to Eq. 14 is that a scalar product value of an adaptive filter weight and a spreading code of 1^th multipath is maintained as 1. There is no condition for an scalar product value of other multipath's code. An output of the adaptive filter has multipath IPI.

[0109] By using Langlangian multiplier, w_kpm,l(n) can be calculated from Eq. 14, wherein w_kpm,l(n)minimizes the object function Eq. 14. It is expressed as following equation. $\begin{array}{cc}\begin{array}{c}{W}_{\mathrm{kpm},l}^0=\frac{R_m^{-1}{s}_{\mathrm{pl}}}{s_{\mathrm{pl}}^H{R}_m^{-1}{s}_{\mathrm{pl}}}+\\ \underset{il}{\underset{i=1}{\overset{L}{}}}\frac{h_{\mathrm{pm},l}^{*}{h}_{\mathrm{pm},i}\left(s_{\mathrm{pl}}^H{R}_m^{-1}{s}_{\mathrm{pl}}{R}_m^{-1}{s}_{\mathrm{pi}}-s_{\mathrm{pl}}^H{R}_m^{-1}{s}_{\mathrm{pi}}{R}_m^{-1}{s}_{\mathrm{pl}}\right)}{s_{\mathrm{pl}}^H{R}_m^{-1}{s}_{\mathrm{pl}}}\end{array}& \mathrm{Eq}.15\end{array}$

[0110] As shown in Eq. 15, w_kpm,l(n) is identical about p, m, l without consideration of a spreading code. That is, Eq. 15 shows w_kpm,l(n) in a fixed channel, which varied very slowly.

[0111] If the channel is changed quickly, then w_kpm,l(n) is expressed as following equation. $\begin{array}{cc}{w}_{\mathrm{kpm},l}^0=\frac{R_m^{-1}{s}_{\mathrm{pl}}}{s_{\mathrm{pl}}^H{R}_m^{-1}{s}_{\mathrm{pl}}}& \mathrm{Eq}.16\end{array}$

[0112] Eq. 16 has optimum filter weight therefore, w⁰_kpm,l(n) can be rewritten as w_kpm,l(n) and excessive mean square error can be reduced corresponding to the number of spreading codes by updating the adaptive filter weight w_kpm,l(n) by adaptively combining the error signals based on data transmitted with spreading code as a reference signal.

[0113] FIG. 5 is a detailed diagram illustrating an adaptive filter block 36 and 46 in FIGS. 3 and 4 in accordance with the present invention.

[0114] Referring to FIG. 5, the adaptive filter block 36 or 46 includes a multi-condition filter weight updating unit 51 and a plurality of adaptive filter bank 52.

[0115] The multi-condition filter weight updating unit 52 updates the adaptive filter weight based on the error vector e_ml(n) from the error generation block 34 or 44 and passed to corresponding adaptive filter bank 52.

[0116] The adaptive filter bank 52 processes w_pm,l(n) from the multi-condition filter updating block 51 and r_m(n) from the input buffer 33 and 43 and generates y_pm,l(n).

[0117] FIG. 6 is a diagram depicting an adaptive filter block in FIGS. 3 and 4 in accordance with another embodiment of the present invention.

[0118] Referring to FIG. 6, the adaptive filter block is identical to the adaptive filter block in FIG. 5 except a one-condition filter weight updating block 61 is operated by one condition weight update algorithm.

[0119] The one condition filter weight updating block 61 is explained in detail as followings.

[0120] The one-condition filter weight updating block 61 updates the adaptive filter weight of each transmitting antenna p(1<p<P) based on below equation. $\begin{array}{cc}{w}_{\mathrm{pm},l}\left(n+1\right)=w_{\mathrm{pm},l}\left(n\right)+P_{s\mathrm{pl}}^{}\left(\underset{k=0}{\overset{K}{}}{e}_{\mathrm{kpm},l}^{*}\left(n\right)r_m\left(n\right)\right)& \mathrm{Eq}.17\end{array}$

[0121] P_{s pl} is defined as $\frac{I-s_{\mathrm{pl}}{s}_{\mathrm{pl}}^H}{\left(s_{\mathrm{pl}}^H{s}_{\mathrm{pl}}\right)}$

[0122] and it is an orthogonal projection matrix which projects all vectors to a space orthogonal to s_pl. If the adaptive filter weight is updated by using Eq. 17, there is a deviation on a channel estimation value as following equation in case of Eq. 12 is used for estimating channel, wherein Eq. 12 is a simple mean channel estimate method. $\begin{array}{cc}E\left[{\hat{h}}_{\mathrm{pm},l}\left(n\right)\right]=h_{\mathrm{pm},l}\left(n\right)+\underset{l^{}1}{\underset{l^{}=1}{\overset{L}{}}}{h}_{\mathrm{pm},l}{w}_{\mathrm{pm},l}^H\left(n\right)S_{\mathrm{pl}}& \mathrm{Eq}.18\end{array}$

[0123] The deviation of the channel estimation value decrease data detection performance.

[0124] For overcoming this problem, a maximum likelihood channel estimation method is used for gaining a biased channel estimation value. The maximum likelihood channel estimation method is expressed as following equation. $\begin{array}{cc}\left[\begin{array}{c}{\hat{h}}_{\mathrm{pm},1}\\ {\hat{h}}_{\mathrm{pm},2}\\ \\ \\ \\ {\hat{h}}_{\mathrm{pm},L}\end{array}\right]=\frac{1}{N_p}\underset{i=1}{\overset{N_p}{}}{\left[\begin{array}{cccc}1& w_{\mathrm{pm},1}^H\left(n-iQ\right)s_{\mathrm{p2}}& ,,& w_{\mathrm{pm},1}^H\left(n-i\right)s_{\mathrm{pL}}\\ w_{\mathrm{pm},2}^H\left(n-i\right)s_{\mathrm{p1}}& 1& ,,& w_{\mathrm{pm},2}^H\left(n-i\right)s_{\mathrm{pL}}\\ & & ,,& \\ & & ,,& \\ & & ,,& \\ w_{\mathrm{pm},L}^H\left(n-i\right)s_{\mathrm{p1}}& w_{\mathrm{pm},L}^H\left(n-i\right)s_{\mathrm{p2}}& ,,& 1\end{array}\right]}^{-1}{d}_{0p}^{*}\left(n-i\right)w_{\mathrm{pm}}^H{r}_m\left(n-i\right)& \mathrm{Eq}.10\end{array}$

[0125] wherein $W_{\mathrm{pm}}^H=\left[\begin{array}{cccc}{w}_{pm,1}& w_{pm,2}& ,,& w_{pm,L}\end{array}\right].$

[0126] FIG. 7 is a detailed diagram showing an adaptive filter bank block in accordance with a preferred embodiment of the present invention.

[0127] Referring to FIG. 7, the adaptive filter bank block includes a plurality of adaptive filters. The adaptive filter receives a standard adaptive filter weight from the filter weight updating block 51 or 61 and sets the standard adaptive filter weight as an adaptive filter weight for a pilot symbol. Also, each spreading code is scrambled and set as an adaptive filter weight for transmission data.

[0128] FIG. 8 is a detailed diagram showing a data decider in accordance with the present invention.

[0129] The data decider decides data symbol closest to a constellation by using a combined signal of Eq. 7 and output a final estimation value of a transmission data symbol.

[0130] FIG. 9 is a graph showing a bit error rate of the present invention with 4 transmitting antennas and 4 receiving antennas in frequency flat fading channel environment.

[0131] Referring to FIG. 9, the present invention has better detection performance than conventional V-BLAST receiver. It is because the receiver of the present invention can use full diversity gain. In case of embodiment A, the performance is getting lower than the conventional V-BLAST receiver since the embodiment A dose not eliminated interference cause by multi-users.

[0132] FIG. 10 is a graph depicting a receiving diversity gain obtained in the present invention.

[0133] Referring to FIG. 10, in case of conventional V-BLAST transmission system, 2.5 order of diversity when four receiving antennas and four transmitting antennas are used.

[0134] As shown in FIG. 10, a characteristic line of the conventional V-BLAST transmission system is located in between a characteristic line of the present invention with three receiving antennas and a characteristic line of the present invention with two receiving antennas. It tells that the present invention obtains order of diversity as many as the number of receiving antennas.

[0135] FIG. 11 is a graph showing a performance of the present invention in three frequency selective fading channel environment.

[0136] Four transmitting antennas and two receiving antennas are used in the present invention. Since there is no V-BLAST system for DS/CDMA system in frequency selective fading channels, the graphs shows performances of the present invention. As shown in FIG. 11, the present invention has excellent data detection performance.

[0137] As mentioned above, the present invention can obtain diversity gain of a theoretical maximum order of receiving diversity.

[0138] The present invention effectively transmits and receives data in the DS/CDMA system equipping the MIMO antenna.

[0139] Also, the present invention has excellent detection performance by receiving data with full order of diversity, which is multiple of the number of receiving antennas and the number of multipath, without complicated operation such as a decision feed equalizer.

[0140] Furthermore, the present invention can eliminate interference between data simultaneously transmitted through different antennas by using various conventional receiving method developed for one transmission antenna with simple modification of the method.

[0141] While the present invention has been described with respect to certain preferred embodiments, it will be apparent to those skilled in the art that various changes and modifications may be made without departing from the scope of the invention as defined in the following claims.