Signal detection by a receiver in a multiple antenna time-dispersive system

• US 7,359,466 B2
• Filed: 08/24/2001
• Issued: 04/15/2008
• Est. Priority Date: 08/24/2001
• Status: Expired due to Term

• Current Assignee: LGS Innovations LLC (CACI International Incorporated)
• Original Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Inventors: Sandell, Magnus, Mailaender, Laurence Eugene, Viswanathan, Harish, Huang, Howard C
• Primary Examiner(s): PHU, PHUONG M
• Application Number:
  US09/938,453
• Publication Number:
  US 20030076908A1
• Time in Patent Office: 2,426 Days
• Field of Search: 375/220, 375229-236, 375/267, 375/350, 375/347, 375/144, 375/346, 375/260, 375/148, 375/349, 370/208, 370/209, 370/210, 370/347, 370/342, 455/92, 714/794
• US Class Current: 375/349
• CPC Class Codes: H04B 1/707 using direct sequence modulation
• CPC Class Codes: H04B 7/0891 Space-time diversity rake receivers H04B1/7115; space-time decoding H04L1/0631
• CPC Class Codes: H04L 1/06 using space diversity
• CPC Class Codes: H04L 2025/03605 Block algorithms
• CPC Class Codes: H04L 2025/03617 Time recursive algorithms H04L2025/03643 takes precedence
• CPC Class Codes: H04L 25/0242 using matrix methods
• CPC Class Codes: H04L 25/03012 operating in the time domain H04L25/03165, H04L25/03178 take precedence
• CPC Class Codes: H04L 25/03331 Arrangements for the joint estimation of multiple sequences

Abstract

In a MIMO system the bit error rate floor caused by time dispersion is reduced by employing a joint minimum mean square error (MMSE) equalizer for all of the respective transmit antenna—receive antenna pairings that are possible in the MIMO system. The resulting joint equalization compensates not only for the impact of the channel on the transmit antenna—receive antenna pairings but also for the interference of the other transmit antennas on any given receive antenna. The joint equalization outperforms simply replicating the prior art minimum mean square error (MMSE) equalizer for each transmit antenna—receive antenna pairings.

Claims

1. A method for compensating for time dispersion in a receiver of a wireless system that has a plurality of transmit antennas and a plurality of a receive antennas, the method comprising the steps of:

receiving samples for each receive antenna;

determining a joint equalizer solution using channel information for at least one pairing of at least one of said transmit antennas and said receive antennas and said received samples of at least two of said receive antennas; and

applying said determined joint equalizer solution to said received samples from at least one of said receive antennas to develop equalized samples;

wherein said step of determining a joint equalizer solution is performed at least partly in the discrete frequency domain.

2. The invention as defined in claim 1 wherein said joint equalizer solution is a joint minimum mean square error (MMSE) solution.

3. The invention as defined in claim 1 further comprising the step of estimating a channel for said at least one pairing of at least one of said transmit antennas and said receive antennas.

4. The invention as defined in claim 1 said step of applying said determined joint equalizer solution is performed in the frequency domain.

6. The invention as defined in claim 1 wherein at least two of said transmit antennas transmit at different rates.

7. The invention as defined in claim 1 wherein at least two of said transmit antennas transmit using different transmit constellations.

8. The invention as defined in claim 1 further comprising the step of performing soft mapping using a version of said equalized samples.

9. The invention as defined in claim 8 wherein said version of said equalized samples are despread samples.

10. The invention as defined in claim 8 wherein said step of performing soft mapping further comprises the step of spatial whitening said version of said equalized samples.

11. The invention as defined in claim 8 herein said step of performing soft mapping further comprises the step of performing a posteriori probability (APP) metric processing an said version of said equalized samples.

12. The invention as defined in claim 1 wherein said determining step is performed multiple times, once for each one of said transmit antennas.

13. The invention as defined in claim 1 wherein said determining and applying steps are iterated multiple times over a symbol period, and iteration for each one of said transmit antennas, and said method further comprises, for each iteration, the steps of:

generating a representation of signals that would have arrived had a particular symbol for a currently being processed transmit antenna had been transmitted; and

subtracting said representation from said samples received for each receive antenna.

14. The invention as defined in claim 1 wherein said joint equalizer solution is one from the group consisting of:

a joint least mean square (LMS) solution, a joint recursive least squares (RLS) solution, or a joint minimum intersymbol interference (ISI) subject to an anchor condition solution.

23. The invention as defined in claim 1 further comprising

an error correction decoder coupled to said soft bit mapper;

a space time regenerator coupled to said error correction decoder; and

a buffer-subtractor coupled between said signal detectors and said joint equalizer and between said space time regenerator and said joint equalizer.

24. The invention as defined in claim 23 further comprising a front end processor coupled to said buffer-subtractor.

5. A method for compensating for time dispersion in a receiver of a wireless system that has a plurality of transmit antennas and a plurality of receive antennas, the method comprising the steps of:

receiving samples for each receive antenna;

determining a joint equalizer solution using channel information for at least one pairing of at least one of said transmit antennas and said receive antennas and said received samples of at least two of said receive antennas;

applying said determined joint equalizer solution to said received samples from at least one of said receive antennas to develop equalized samples; and

despreading said equalized samples.

15. A receiver for use in a multiple-input multiple-output (MIMO) system in which a plurality of signal detectors receive signals transmitted by a plurality of signal sources, said receiver comprising:

a joint equalizer that develops a joint equalizer solution using channel information for at least one pairing of at least one of said signal sources and said signal detectors and received samples of at least two of said signal detectors and supplies as an output a signal that includes at least said joint equalizer solution applied to a signal received by at least one of said signal detectors;

a soft bit mapper for developing soft bits from said joint equalizer output; and

a despreader interposed between said joint equalizer and said soft bit mapper.

16. The invention as defined in claim 15 wherein said joint equalizer solution is a joint minimum mean square error (MMSE) equalizer solution.

17. The invention as defined in claim 15 wherein said soft bit mapper further comprises an a posteriori probability (APP) metric processor.

18. The invention as defined in claim 15 wherein said soft bit mapper further comprises a spatial whitening unit.

19. The invention as defined in claim 15 where at least two of said transmit sources transmits signals at different rates.

20. The invention as defined in claim 15 wherein at least two of said transmit sources transmits signals using different transmit constellations.

21. The invention as defined in claim 15 further comprising:

a space time regenerator coupled to said joint equalizer; and

a buffer-subtractor coupled between said signal detectors and said joint equalizer and between said space time regenerator and said joint equalizer.

22. The invention as defined in claim 21 further comprising a front end processor coupled to said buffer-subtractor.

25. The invention as defined in claim 15 further comprising an order controller for determining an order in which signals from said signal detectors will be processed by said joint equalizer.

31. The invention as defined in claim 15 wherein said MIMO system is a wireless system, said signal sources are transmit antennas and said detectors are antennas of said receiver.

32. The invention as defined in claim 15 wherein said joint equalizer develops said joint equalizer solution as a function of estimates of the channels between each of said signal sources and said signal detectors.

33. The invention as defined in claim 15 wherein said joint equalizer develops and applies said joint equalizer solution in a time domain.

34. The invention as defined in claim 15 wherein said joint equalizer develops and applies said joint equalizer solution in a frequency domain.

35. The invention as defined in claim 16 wherein said joint equalizer develops and applies said joint equalizer solution in a discrete frequency domain and applies said joint minimum mean square error (MMSE) equalizer solution in a time domain.

36. The invention as defined in claim 16 wherein said joint equalizer develops said joint minimum mean square error (MMSE) equalizer solution by computing

$W=A$

⁢

( Γ

⁢

( H ) ) H ⁢

( ( Γ

⁢

( H ) ) H ⁢

Γ

⁢

( H ) + σ

n 2 σ

x 2 ⁢

R pp ) - 1

where

Γ

(H) is a MIMO convolution operator,

X^H means the Hermitian transpose of X, which is the complex conjugate transpose of the vector or matrix X,

A is a delay matrix,

σ

_n² is the background noise plus interference power,

σ

_x² is the sum of the power received by all said signal detectors from all of said signal sources,

noise covariance R_pp,

and said joint equalizer applies said joint minimum mean square error (MMSE) equalizer solution by computing

y(k)=Wr(k)

where r(k) is a vector of received samples at time k, and

is y(k) is the vector output at time k representing the application of said joint minimum mean square error (MMSE) equalizer to said vector of received samples at time k.

37. The invention as defined in claim 15 wherein said joint equalizer solution is one from the group consisting of:

a joint least mean square (LMS) solution, a joint recursive least squares (RLS) solution, or a joint minimum intersymbol interference (ISI) subject to an anchor condition solution.

26. A receiver for use in a multiple-input multiple-output (MIMO) system in which a plurality of signal detectors receive signals transmitted by a plurality of signal sources, said receiver comprising:

a joint equalizer that develops a joint equalizer solution using channel information for at least one pairing of said at least one of said signal sources and said signal detectors and received samples of at least two of said signal detectors and supplies as an output a signal that includes at least said equalizer solution applied to a signal received by at least one of said signal detectors; and

a soft bit mapper for developing soft bits from said joint equalizer output;

wherein said joint equalizer solution is a joint minimum mean square error (MMSE) equalizer solution and wherein said join equalizer further comprises;

a first plurality of fast Fourier transform processors, each of said fast Fourier transform processors being coupled to receive a respective input corresponding to a signal received by one of said signal detectors and supplying as an output a discrete frequency domain representation thereof;

a plurality of channel estimation units each of which is coupled to receive a is respective input corresponding to a signal received by one of said signal detectors which develops a channel estimate for each channel between each respective signal source and each respective signal detector;

a second plurality of fast Fourier transform processors, each of said second plurality of fast Fourier transform processors coupled to receive a respective input corresponding to a channel estimate for a respective one of said channels between said signal sources and said signal detectors and supplying as an output a discrete frequency domain representation thereof;

an MMSE detection per frequency bin processor coupled to receive as inputs said outputs from said first plurality of fast Fourier transform processors and from said second plurality of fast Fourier transform processors to produce a discrete frequency domain representation of an application of said joint minimum mean square error (MMSE) equalizer solution to said signals received by each of said signal detectors; and

a plurality of inverse fast Fourier transform processors which convert said discrete frequency domain representation of an application of said joint minimum mean square error (MMSE) equalizer solution to the time domain.

27. The invention as defined in claim 26 wherein M is the number of signal sources and N is the number of signal detectors, and

wherein said MMSE detection per frequency bin processor performs the equalization in the frequency domain by computing

z(ω

)=(H(ω

)^HH(ω

)+σ

²I)<sup>−

1</sup>H(ω

)^Hr(ω

)

where

$H<br><br>\left(\mathrm{\omega <br><br>}\right)=\left[\begin{array}{ccc}{h}_{1,1}<br><br>\left(\mathrm{\omega <br><br>}\right)& \mathrm{\cdots <br><br>}& h_{1,M}<br><br>\left(\mathrm{\omega <br><br>}\right)\\ \mathrm{⋮<br><br>}& \mathrm{⋰<br><br>}& \mathrm{⋮<br><br>}\\ h_{N,1}<br><br>\left(\mathrm{\omega <br><br>}\right)& \mathrm{\cdots <br><br>}& h_{N,M}<br><br>\left(\mathrm{\omega <br><br>}\right)\end{array}\right]$ $r<br><br>\left(\mathrm{\omega <br><br>}\right)=\left[\begin{array}{c}{r}_1<br><br>\left(\mathrm{\omega <br><br>}\right)\\ r_2<br><br>\left(\mathrm{\omega <br><br>}\right)\\ \mathrm{⋮<br><br>}\\ r_N<br><br>\left(\mathrm{\omega <br><br>}\right)\end{array}\right]$ ${\mathrm{\sigma <br><br>}}^2=\frac{{\mathrm{\sigma <br><br>}}_n^2}{{\mathrm{\sigma <br><br>}}_x^2}$

σ

² is the background noise plus interference power,

σ

_x² is the sum of the power received by all said signal detectors from all of said signal sources,

each r(ω

) is said output of a one of said first plurality of first Fourier transform processors,

each h(ω

) is said output of a one of said second plurality of first Fourier transform processors,

I is the identity matrix,

X^H means the Hermitian transpose of X, which is the complex conjugate transpose of the vector or matrix X, and

z(ω

) is the equalized output vector in the discrete frequency domain for frequency ω

.

28. A receiver for use in a multiple-input multiple-output (MIMO) system in which a plurality of signal detectors receive signals transmitted by a plurality of signal sources, said receiver comprising:

a joint a equalizer that develops a joint equalizer solution using channel information for at least one pairing of said at least one of said signal sources and said signal detectors and received samples of at least two of said signal detectors and supplies as an output a signal that includes at least said equalizer solution applied to a signal received by at least one of said signal detectors; and

a soft bit mapper for developing soft bits from said joint equalizer output;

wherein said joint equalizer solution is a joint minimum mean square error (MMSE) equalizer solution and wherein said joint equalizer further comprises;

a plurality of channel estimation units each of which coupled to receive a respective input corresponding to a signal received by one of said signal detectors which develops a channel estimate for each channel between each respective signal source and each respective signal detector;

a plurality of fast Fourier transform processors, each of said plurality of fast Fourier transform processors being coupled to receive a respective input corresponding to a channel estimate for a respective one of said channels between said signal sources and said signal detectors and supplying as an output a discrete frequency domain representation thereof;

an MMSE tap weight calculator coupled to receive as inputs said outputs from said plurality of fast Fourier transform processors to produce a discrete frequency domain representation of a joint minimum mean square error (MMSE) equalizer solution to said signals received by each of said signal detectors;

a plurality of inverse fast Fourier transform processors which convert said discrete frequency domain representation of a joint minimum mean square error (MMSE) equalizer solution to matrices of filter coefficients in the time domain; and

a matrix finite impulse response (FIR) fitter coupled to apply said matrices of filter coefficients in the time domain to said signals received by said signal detectors.

29. The invention as defined in claim 28 wherein M is the number of signal sources and N is the number of signal detectors, and wherein said MMSE equalizer solution is developed by computing:

S(ω

)=(H(ω

)^HH(ω

)+σ

²I)<sup>−

1</sup>H(ω

)^H

where

$H<br><br>\left(\mathrm{\omega <br><br>}\right)=\left[\begin{array}{ccc}{h}_{1,1}<br><br>\left(\mathrm{\omega <br><br>}\right)& \mathrm{\cdots <br><br>}& h_{1,M}<br><br>\left(\mathrm{\omega <br><br>}\right)\\ \mathrm{⋮<br><br>}& \mathrm{⋰<br><br>}& \mathrm{⋮<br><br>}\\ h_{N,1}<br><br>\left(\mathrm{\omega <br><br>}\right)& \mathrm{\cdots <br><br>}& h_{N,M}<br><br>\left(\mathrm{\omega <br><br>}\right)\end{array}\right]$ ${\mathrm{\sigma <br><br>}}^2=\frac{{\mathrm{\sigma <br><br>}}_n^2}{{\mathrm{\sigma <br><br>}}_x^2}$

σ

_n² is the background noise plus interference power,

σ

_x² is the sum of the power received by all said signal detectors from all of said signal sources,

each h(ω

) is said output of a one of said plurality of fast Fourier transform processors,

I is the identity matrix,

X^H means the Hermitian transpose of X, which is the complex conjugate transpose of the vector or matrix X, and

S(ω

) is a frequency domain matrix representing the equalization for frequency ω

.

30. The invention as defined in claim 29 wherein said matrix FIR filter applies said matrices of filter coefficients in the time domain to said signals received by said signal detectors by computing

$y$

( k ) = ∑

j = 0 F - 1 ⁢

S j ⁢

r ⁢

( k - j )

where y(k) is a vector output at time k, y having one component for each of said signal sources,

S_j is a M×

N filter matrix for delay j, which is the inverse Fourier transform of S(ω

),

r(k) is a vector of received samples at time k, and

F is the number of samples taken for each fast Fourier transform by each of said plurality of fast Fourier transform processors.

38. A receiver for use in a multiple-input multiple-output (MIMO) system in which a plurality of receive antennas receive signals transmitted by a plurality of transmit antennas, said receiver comprising:

means for (i) developing a joint equalizer solution using channel information for at least one pairing of at least one of said transmit antennas and said receive antennas and said received samples of at least two of said receive antennas, said joint equalizer solution being developed at least partly in a frequency domain, and (ii) supplying as an output a signal that includes at least said joint equalizer solution applied to signal received by at least one of said receive antennas; and

means for developing soft bits from said output signal.