AU2006201932A1 - Efficient Multiple Input Multiple Output System for Multi-Path Fading Channels - Google Patents

Description

P001 Section 29 Regulation 3.2(2)

AUSTRALIA

Patents Act 1990 COMPLETE SPECIFICATION STANDARD PATENT Application Number: Lodged: Invention Title: Efficient Multiple Input Multiple Output System for Multi-Path Fading Channels The following statement is a full description of this invention, including the best method of performing it known to us: EFFICIENT MULTIPLE INPUT MULTIPLE OUTPUT SYSTEM FOR MULTI- PATH FADING CHANNELS FIELD OF INVENTION This invention relates generally to wireless communication systems. In particular, the invention relates to transferring signals using antenna arrays.

BACKGROUND

A multiple input multiple output (MIMO) system is described in Figure 1.

Multiple transmission and/or reception antennas 121 to 12 M 161 to 16 N (16) are used to transfer the communication. Each antenna 12, 16 is spatially separated from the other antennas 12, 16. A transmitter 10 using its antenna array 12 transmits a communication to a receiver 18 through a wireless air interface 18. The receiver 18 receives the communication using its antenna array 16. Using both multiple transmission and reception antennas 12, 16 is referred to as multiple input multiple output (MIMO) processing.

Typically, MIMO processing employs multiple antennas at both the base station transmitter and user equipment receiver. While the deployment of base station antenna arrays is already commonly used in wireless communication systems, the simultaneous deployment of base station and user equipment arrays enable significant increases in capacity and data rates by opening up multiple signalling dimensions.

Available MIMO algorithms address a single-path fading channels.

However, wireless communication systems are characterised by multipath fading channels. Algorithms that are designed for single-path fading channels, typically exhibit severe degradation in presence of multipath.

Accordingly, it is desirable to have other MIMO systems.

Any discussion of documents, devices, acts or knowledge in this specification is included to explain the context' of the invention. It should not be taken as an admission that any of the material formed part of the prior art base or the common general knowledge in the relevant art in Australia on or before the priority date of the claims herein.

SUMMARY

In accordance with the present invention, there is provided a base station including: a plurality of receiving antennas for receiving signals from a plurality of transmitting antennas at a single site, wherein the signals are encoded by a multicode vector encoder; a plurality of demodulating devices, respectively coupled to the plurality of receiving antennas, for demodulating the received signals; a plurality of sampling devices, respectively coupled to the plurality of demodulating devices, each of the sampling devices for sampling a received version of the received signals as received by one of the plurality of receiving antennas to produce a combined received signal; a channel estimation device, coupled to the plurality of sampling devices, for determining a channel response for each receiving and transmitting antenna combination and producing an overall channel response; a multiple input multiple output (MIMO) channel equalization device, coupled to the plurality of sampling devices and the channel estimation device, for processing the combined received signals and the overall channel response to produce a spread data vector; and a despreading device for despreading the spread data vector to recover data of the received signals.

Further preferred aspects of this invention may be as defined in the dependent claims 2 to 5 annexed hereto, which claims are hereby made part of the disclosure of this invention.

Comprises/comprising and grammatical variations thereof when used in this specification are to be taken to specify the presence of stated features, integers, steps or components or groups thereof, but do not preclude the presence or addition of one or more other features, integers, steps, components or groups thereof.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is an illustration of a transmitter and a receiver using multiple antennas.

Figure 2 is a simplified block diagram of a preferred MIMO transmitter and receiver.

Figure 3A is an embodiment of a MIMO channel equalisation device.

Figure 3B is a flow chart of a MIMO channel equalisation embodiment.

Figure 4A is an alternate embodiment of a MIMO channel equalisation device.

Figure 4B is a flow chart of a MIMO channel equalisation alternate embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Figure 2 is a simplified block diagram of a multiple input multiple output (MIMO) transmitter and receiver system. The transmitter 20 can be used in a user equipment, a base station or both and the receiver 22 may be used in a base station, a user equipment or both. The MIMO system preferably uses a code division multiple access (CDMA) air interface as shown in Figure 2, such as a frequency division duplex (FDD)/CDMA, time division duplex (TDD)/CDMA or time division synchronous code division multiple access (TD-SCDMA) air interface, although other air interfaces may be used.

[0020] A data vector d is to be transmitted through the wireless air interface by the transmitter 20. The transmitter 20 has M antennas 341 to 3 4

M

(34) in the antenna array. If transmission spatial diversity is not used, Mis one, (a single antenna). The antennas 34 are spatially separated so that a low correlation between their received signals is obtained. For use at a base station transmitter having an angle spread in the range of 1 to 10 degrees, the antenna separation is preferably several wavelengths, such as four wavelengths. For use at a UE receiver, since the angle spread tends to be large, a smaller separation may be used, such as half a wavelength. Based on the implementation, the spatial separations may have other values.

[0021] For a preferred implementation of transferring a multicode transmission as illustrated in Figure 2, the data vector d to be transmitted by the M antennas 34 is encoded by a multicode vector encoder 26. For each of the Q spreading codes Ci to CQ, the data is split into M separate data streams dl,1 to dMQ prior to spreading. The total number of produced data streams is M Q. To illustrate for Ci, data streams dii to dml are produced. Each of the M streams is associated with an antenna 34.

[0022] For each code, the data streams are spread by their code using a corresponding spreading device 281 to 28Q such as a mixer. The spread data streams associated with the same antenna 34 are input into a combiner 3 01 to such as an adder, associated with that antenna 34 of the M antennas 34, producing M spread data vectors, si to SM. Each combined spread data vector, si to SM, is modulated to radio frequency by a modulator 321 to 3 2 M (32) and radiated by its associated antenna 34 through the wireless air interface 24.

[0023] The preferred multicode receiver, implementation as shown in Figure 2 is for use when all of the multicode transmissions experience the same channel response with respect to a transmission and reception antenna pair.

This typically occurs in the downlink. Alternately in the uplink, the receiver 22 of Figure 2 can be used to process a single user's transmissions, when multiple users are transmitting. The other user transmissions are treated as noise.

[0024] At the receiver 22, each transmission antenna's radiated signal is received by each of the N reception antennas 361 to 3 6 N, as a combined received signal. If reception spatial diversity is not used, N is one, (a single antenna). N is preferably equal to or greater than M. Each received antenna's signal is demodulated to baseband by a demodulator 381 to 3 8 N Each demodulated signal is sampled, such as at the chip rate or a multiple of the chip rate, by a sampling device 4 01 to 40N to produce a received vector for each antenna 36, r_ to rN. The combined received vector r comprises ri to rN.

[0025] The combined received vector r is input into a MIMO channel equalization device 44. A training sequence signal r' is input into a channel estimation device 44. The channel estimation device 42 estimates the channel response for each reception and transmission antenna combination. For an i th receiving antenna 36 and ajh transmitting antenna 34, the channel response as a kth instant in time is hij(k). The overall channel response for all antenna combinations at the kth instant of time is per Equation 1A.

hH hl,m (k) hN hN,M (k) Equation 1A [0026] The overall channel response is per Equation lB.

H(O) 0 0 0 0 0 0 0 0 0 H(1) H(0) 0 0 0 0 0 0 0 H(2) H(1) H(0) 0 0 0 0 0 0 SH(L-1) H(L-2) H(L-3) H(0) 0 0 0 0 0 H(L-1) H(0) 0 0 0 0 H(L-1) H(1) 0 0 0 0 0 0 0 H(L-1) Equation 1B [0027] The overall channel response H is passed to the MIMO channel equalization device 44. The MIMO channel equalization device 44 uses the channel response matrix H and equalizes the received vector r to compensate for the channel distortion experienced through the wireless air interface 24, producing a spread data vector s. The spread data vector s is reordered by a spread vector reordering device 46 so that each transmitting antenna's spread data vector si to SM is recovered. Each transmitting antenna's spread data vector si to SM is despread by a despreading device 48 using the spreading codes, Ci to CQ, to estimate data for each of the M encoded data streams for each antenna, dii to A despread data stream decoder 50 combines the data streams dii to dM,Q to recover the original data vector d.

[0028] Each communication transferred by a particular transmission/reception antenna combination experiences a different multi-path environment than the other transmission/reception antenna combinations, due to the spatial diversity. By processing the received multi-path components of all of the reception antennas 3 6 1 to 3 6 N, the capacity and maximum data rate of the system is significantly increased.

[0029] Figure 3A is an embodiment of a MIMO channel equalization device 44 and Figure 3B is a flow chart of a MIMO channel equalization embodiment.

Other embodiments for a MIMO channel equalization device, such as Cholesky or approximate Cholesky decomposition, can also be used. The received samples of the channel impulse responses for each antenna pair are per Equation 2.

hi(k) where k 0, ,L -1 Equation 2 [0030] i is the i th receiving antenna. j is the jth transmitting antenna, k is the k t h sample of the impulse response of length L. The channel impulse response of all reception and transmission antenna pairings for the kth sample is per Equation 3.

h 11 hl,u(k) H(k) hN,l hN,M(k)J Equation 3 [0031] The transmitted spread data vector s has N, M dimension vectors Ns is the number of transmitted data symbols. The overall received vector r has Ns L -2 received vectors of dimension N and is per Equation 4.

H(0) 0 0 0 0 0 0 0 0 H(1) H(O) 0 0 0 0 0 0 0 r(0) H(2) H(1) H(0) 0 0 0 0 0 0 r(1) s(2) r) H(L-1) H(L-2) H(L-3) H(0) 0 0 s(L 1) 0 0 0 H(L-1) H(0) N,+L 0 0 0 0 H(L-1) H(1) s(N, -1) 0 0 0 0 0 0 0 H(L-1) Equation 4 [0032] w is the noise vector. Alternately, Equation 4 can be written as Equation r Hs+w Equation The r and w vectors have a length of L N. The s vector has a length of N,M and His a by N, -M matrix.

[0033] As shown in Equation 4, the H matrix is approximately block circulant. To make the H matrix more block circulant, L-1 block columns are added to the H matrix, to produce an extended H matrix, and a corresponding number of zeros added to the s vector, to produce an extended s vector. The L-1 column blocks are added in accordance with the H matrice's block circulant structure. After the extending of the H matrix and the s vector, the extended H matrix has a dimension of N by M and s has a length of [0034] For shorthand, Ns L -1 is represented by D, so that D Ns +L -1.

The extended H matrix of size DN by DM with blocks of size N by M is decomposed per Equation 6.

M)FM)

Equation 6 F(N) is a block Fourier transform with a block size of N by N and FM is a block Fourier transform with a block size of Mby M. F(N) is per Equation 7.

F 0 IN Equation 7 is the Kronecker product operation and IN is an Nby N identity matrix.

[0035] F(M) is per Equation 8.

F(M F IM Equation 8 IM is an Mby Midentity matrix.

[0036] The block-diagonal matrix preferably, a block-Fourier transform of the first block-column of H, although another column may be used, (after being permuted), step 84. A block-Fourier transform device 62 produces A(N, by taking a block-transform of a column of H. is preferably derived from Equation 9.

diag(N,M) F(NH(,M) M) Equation 9 diag(,m is the block diagonal of M) represents the first block column of width M. By using a single block column of Hto derive the H matrix is approximated as being a block circulant matrix.

[0037] By substituting Equation 6 into Equation 2, Equation 10 results.

A(N,M)F(M) F(N)r Equation [0038] To solve for s, initially, a vector x is determined per Equation 11, step 86.

x= F, Equation 11 x is preferably determined by an Nnon-block discrete Fourier transform of length D. A block-Fourier transform device 64 produces x by taking the block-transform of r.

[0039] Subsequently, a vector y is determined per Equation 12, step 88.

A(N,M) x Equation 12 A y determining device 66 produces y.

[0040] Since A(N,M) is a block diagonal matrix, y is preferably determined on a block-by-block basis by solving D systems of equations of smaller size, such as per Equation 13.

=i Equation 13 AN,M) is the i t h block of yi is the ith Mby 1 sub-vector of y. xi is the ith N by 1 sub-vector of x.

[0041] Since AN,,M) is unstructured, one approach to solve Equation 13 is to use Cholesky decomposition of Equation 14 and forward and backward substitution, although other approaches may be used.

V'\I(NM) )HA! AM) Equation 14 [0042] If the number of receiving antennas N is equal to the number of transmitting antennas, is a square matrix and y can be determined by inverting M. For small values ofN, a matrix inversion may be more efficient than performing a Cholesky decomposition.

[0043] The s vector is determined per Equation 15, step F= Y Equation A block inverse Fourier transform device 68 is used to produce s. One approach to determining s using Equation 15 is by performing Mnon-block inverse discrete Fourier transforms of dimension D, although other approaches may be used.

[0044] Figure 4A is an alternate embodiment of a MIMO channel equalization device 44B and Figure 4B is a flow chart of the MIMO channel equalization alternate embodiment. To determine s, both sides of Equation 2 are multiplied by H, per Equation 16.

HH r= Rs+H H w=Rs+n Equation 16 is the conjugate transpose operation. n is the equalized noise vector. R is the channel cross correlation matrix and is determined by an R determination device using the H matrix, step 92. R for a zero forcing solution is per Equation 17.

R H H Equation 17 [0045] For a minimum mean square errors (MMSE) solution, R is per Equation 18.

R HHH 2

I

Equation 18 a 2 is the variance of the noise vector w and I is an identity matrix.

[0046] The channel cross correlation matrix R has a structure per Equation 19.

R

o RI R 2

R

3 RL-I 0 0 0 0 0 0 RI Ro RI R2 R3 RL- 0 0 0 0 0

R

2 R Ro R 1

R

2

R

3 RL- 0 0 0 0

R

3

R

2

R

1 Ro R R 2 R3 RL-I 0 0 0 R3 R 2 Rf Ro R 1

R

2

R

3 RL- 0 0 RH R R Ro RI R R

R

-I 0

R=

0 R R R Ro Ri R2 R 3 R L

R

0 0 R R 2 R R" Ro R R 2 R3 0 0 0 R H

R

2 R Rf Ro R Rz R.

0 0 0 0 R 1

R

3 Ra R, Ro Ri R2 0 0 00 R, R R Ro R, 0 0 0 0 0 0 RH- R3 R2 Rf Ro Equation 19 [0047] After the adding of L-1 columns to the H matrix as previously described, a close to block circulant approximation of the R matrix can be derived, referred to as the extended R matrix. The dimensions of the extended R matrix are DMby DM.

[0048] Using the extended R matrix, Equation 20 is used to determine s.

H r=Rs Equation [0049] By approximating R and HR as block circulant matrices, R is decomposed per Equation 21.

R F()A (MM)F(M) Equation 21 AM,M) ,is preferably derived from a first block column of R per Equation 22, although another column may be used, (after being permutated), step 94.

diag

M)

Equation 22 diag(MM)(A is the block diagonal of AR(M). A block-Fourier transform device 72 is used to determine AMM) by taking a block-Fourier transform of a block column of R.

[0050] H" is decomposed per Equation 23.

H H AHF F (M,N)F Equation 23 [0051] is preferably determined using the first block column of H per Equation 24, although another block column may be used, (after being permuted), step 96.

diagM,) FM)HM,N) N) Equation 24 diag(MN)(AM,N)) is the block diagonal of A"M,N).

[0052] A block-Fourier transform device 74 is used to determine AM,,v by taking a block-Fourier transform of a block-column of H

H

[0053] By substituting 21 and 23 into 20, Equation 25 results.

AIM,N)F(N) AM,M)F(M) S Equation [0054] Solving for s Equation 26 results.

A(MN)FN) r Equation 26 [0055] Determining s is preferably performed in a four step process.

Initially, x is determined per Equation 27, step 98.

X Fr Equation 27 The block-Fourier transform is preferably performed by an N non-block Fourier transform of length D. A block-Fourier transform device 76 takes a block-Fourier transform of r to determine x.

[0056] In a second step, y is determined per Equation 28, step 100.

AMN)X

Equation 28 A y determining device 78 uses A(MN) and x to determine y.

[0057] In a third step, z is determined per Equation 29, step 102.

M zM) Y Equation 29 A z determining device 80 determines z using AMM) and y.

[0058] Preferably, since A(MM) is a block diagonal matrix, Equation 29 is solved using D systems of smaller size per Equation i=l,.,D Equation [0059] is the t^ block of z is the ih Mby 1 sub-vector of z.

y. is the i th Mby 1 sub-vector of y.

[0060] Since is unstructured, one approach to determine (M~AM)) is to use Cholesky decomposition of (AMM and forward and backward substitution, although other approaches may be used.

[0061] In the fourth step, s is determined per Equation 31 by performing an M non-block inverse discrete Fourier transform of dimension D, step 104.

S Fm)Z Equation 31 A block-inverse Fourier transform device 82 is used to take an inverse block transform of z to produce s.

[0062] To recover data from the estimated spread data vector s using either embodiment, the spread data vector s is separated into its M transmitted data streams, sm where m 1, M, by the spread vector reordering device. The -12spread data vector s is the result of concatenating the data stream vectors sm and reordering by grouping the same chip intervals together, as per Equation 32.

SM,

S=

SI,N,

S

M,N, J Equation 32 smj demotes the jth chip interval of the m 1 t data stream.

[0063] To recover each spread data vector stream sm, the chips of the estimated spread data vector s are reordered per Equation 33, producing

SREORDERED.

Si' '1,

S

,Nc S1 REODERED

S

M,1 SM

SM.N,

Equation 33 [0064] Each of these data stream spread data vectors Sm is despread by the corresponding spreading codes, Ci to CQ, to estimate the symbols of that data stream by the despreading device 48. The estimated symbols for each data stream are decoded by the despread data stream decoder 50 to recover the original data d.

-13-

Claims (6)

1. A base station including: a plurality of receiving antennas for receiving signals from a plurality of transmitting antennas at a single site, wherein the signals are encoded by a multi- code vector encoder; a plurality of demodulating devices, respectively coupled to the plurality of receiving antennas, for demodulating the received signals; a plurality of sampling devices, respectively coupled to the plurality of demodulating devices, each of the sampling devices for sampling a received version of the received signals as received by one of the plurality of receiving antennas to produce a combined received signal; a channel estimation device, coupled to the plurality of sampling devices, for determining a channel response for each receiving and transmitting antenna combination and producing an overall channel response; a multiple input multiple output (MIMO) channel equalization device, coupled to the plurality of sampling devices and the channel estimation device, for processing the combined received signals and the overall channel response to produce a spread data vector; and a despreading device for despreading the spread data vector to recover data of the received signals.

2. The base station of claim 1 wherein the MIMO channel equalization device including: a block-Fourier transform device for processing a block-column of the overall channel response to produce a diagonal matrix; a block Fourier transform device for processing the combined received signal to take a block-Fourier transform of the combined received signal; a processing device for using the diagonal matrix and the block-Fourier transform of the combined received signal to produce a Fourier transform of the spread data vector; and an inverse block-Fourier transform device for taking an inverse block Fourier transform of the Fourier transform of the spread data vector.

3. The base station of claim 1 further including a spread vector reordering device for reordering the spread data vector so that chips transmitted by each transmitting antenna are grouped together.

4. The base station of claim 1 further including a despread data stream decoder for decoding data recovered by the despreading device.

The base station of claim 1 wherein the MIMO channel equalization device including: a cross channel correlation determining device for producing a cross channel correlation matrix using the overall channel response; a block-Fourier transform device for processing a block-column of the cross channel correlation matrix to produce a cross correlation diagonal matrix; a block-Fourier transform device for processing a block-column of the overall channel response to produce a channel response diagonal matrix; a block-Fourier transform device for processing the combined received signal to take a block-Fourier transform of the combined received signal; a combining device for combining the channel response diagonal matrix and the block-Fourier transform of the combined received signal to produce a combined result; a processing device for using the cross correlation diagonal matrix and the combined result to produce a Fourier transform of the spread data vector; and a block-inverse Fourier transform device for processing the Fourier transform of the spread data vector to produce the spread data vector.

6. A base station substantially as herein described with reference to the accompanying drawings (excluding Figure 1). DATED this 8 th day of May 2006 INTERDIGITAL COMMUNICATIONS CORPORATION WATERMARK PATENT TRADE MARK ATTORNEYS 290 BURWOOD ROAD HAWTHORN VICTORIA 3122 AUSTRALIA P23954AU01