System and method for wireless communication systems

• US 20040190636A1
• Filed: 03/31/2003
• Published: 09/30/2004
• Est. Priority Date: 03/31/2003
• Status: Active Grant

• Current Assignee: Vecima Networks Incorporated
• Original Assignee: Vecima Networks Incorporated
• Inventors: Oprea, Alexandru M.
• Granted Patent: US 7,327,795 B2
• Time in Patent Office: Days
• US Class Current: 375/260
• CPC Class Codes: H04B 7/0417 Feedback systems
• CPC Class Codes: H04B 7/0443 utilizing "waterfilling" technique
• CPC Class Codes: H04B 7/0626 Channel coefficients, e.g. channel state information [CSI]
• CPC Class Codes: H04B 7/0669 using different channel coding between antennas space-time coding H04L1/0618
• CPC Class Codes: H04B 7/0848 Joint weighting
• CPC Class Codes: H04L 1/0003 by switching between different modulation schemes
• CPC Class Codes: H04L 1/0026 Transmission of channel quality indication
• CPC Class Codes: H04L 1/0643 block codes
• CPC Class Codes: H04L 27/2601 Multicarrier modulation systems

Abstract

A system and method for transmitting a plurality of input data symbol sub-streams over a plurality of spatial-subspace channels of a sub-carrier between a transmitter and the receiver. The plurality of input data symbol sub-streams are partitioned in a plurality of super-frames of data and weighted by a weight matrix derived from the singular value decomposition of a channel matrix corresponding to the sub-carrier by applying a partial SVD algorithm. The transmitter further inserts subspace training symbols into the plurality of input data symbol sub-streams and the receiver periodically processes the sub-space training symbols during each super-frame of the plurality of super-frames for estimating output data related to the input data symbol stream.

Claims

1. A transmitter for processing an input data symbol stream for transmission over a plurality of spatial-subspace channels of a sub-carrier, wherein said transmitter comprises:

a) a transmitter SVD unit for calculating an approximation of a transmit weight matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm in which said transmitter SVD unit truncates an interim version of said transmit weight matrix to calculate said approximation of said transmit weight matrix; and

,

b) a transmitter weighting unit connected to said transmitter SVD unit for weighting a plurality of input data symbol sub-streams with said approximation of said transmit weight matrix for distributing said plurality of input data symbol sub-streams along said plurality of spatial-subspace channels, said plurality of input data symbol sub-streams being derived from said input data symbol stream.

2. The transmitter of claim 1, wherein said transmitter SVD unit performs said partial SVD algorithm by performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, said interim weight matrix V_k′

being said interim version of said transmit weight matrix, wherein said transmitter SVD unit further determines a number p of said plurality of spatial-subspace channels, and truncates and zero-pads said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

3. The transmitter of claim 2, wherein said transmitter SVD unit further performs n2 iterations of said iterative SVD algorithm on said truncated matrices U_k″

, Λ

_k″ and

V_k″

to obtain an approximate singular value matrix Λ

_k and said approximation of said transmit weight matrix V_k.

4. The transmitter of claim 3, wherein n1 and n2 are constant integers.

5. The transmitter of claim 4, wherein said constant integers are chosen from the range comprising integers 1 to 4.

6. The transmitter of claim 2, wherein n1 is a variable integer bounded by a constant nmax1, and n1 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said interim matrix Λ

_k′

is less than or equal to a pre-specified fraction of a sum of squared magnitudes of diagonal components of said interim matrix Λ

_k′

.

7. The transmitter of claim 3, wherein n2 is a variable integer bounded by a constant nmax2, and n2 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said truncated matrix Λ

_k″

is less than or equal to a pre-specified fraction of a sum of squared magnitudes of diagonal components of said truncated matrix Λ

_k″

.

8. The transmitter of claim 3, wherein said transmitter SVD unit further applies the partial SVD algorithm to said approximate singular value matrix Λ

_k.

9. A receiver for processing received spatial-subspace data on a plurality of spatial-subspace channels of a sub-carrier, wherein said receiver comprises:

a) a receiver SVD unit for calculating an approximation of a receive weight matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm in which said receiver SVD unit truncates an interim version of said receive weight matrix to calculate said approximation of said receive weight matrix; and

,

b) a receiver weighting unit connected to said SVD unit for weighting said received spatial-subspace data with said approximation of said receive weight matrix.

10. The receiver of claim 9, wherein said receiver SVD unit performs said partial SVD algorithm by performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, a complex conjugate transpose of said interim weight matrix U_k′

being said interim version of said receive weight matrix, wherein said receiver SVD unit further determines a number p of said plurality of spatial-subspace channels and truncates and zero-pads said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

11. The receiver of claim 10, wherein said receiver SVD unit further performs n2 iterations of said iterative SVD algorithm on said truncated matrices Λ

_k″

, U_k″ and

V_k″

to obtain an approximate singular value matrix Λ

_k and said approximation of said receive weight matrix U_k*.

12. The receiver of claim 11, wherein n1 and n2 are constant integers.

13. The receiver of claim 12, wherein said constant integers are chosen from the range comprising integers 1 to 4.

14. The receiver of claim 10, wherein n1 is a variable integer bounded by a constant nmax1, and n1 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said interim matrix Λ

_k′

is less than a pre-specified fraction of a sum of squared magnitudes of diagonal components of said interim matrix Λ

_k′

.

15. The receiver of claim 11, wherein n2 is a variable integer bounded by a constant nmax2, and n2 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said truncated matrix Λ

_k″

is less than a pre-specified fraction of a sum of squared magnitudes of diagonal components of said truncated matrix Λ

_k″

.

16. The receiver of claim 11, wherein said receiver SVD unit further applies the partial SVD algorithm to said approximate singular value matrix Λ

_k.

17. A method for processing an input data symbol stream for transmission over a plurality of spatial-subspace channels of a sub-carrier, wherein said method comprises:

a) calculating an approximation of a transmit weight matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm which comprises truncating an interim version of said transmit weight matrix to calculate said approximation of said transmit weight matrix; and

,

b) weighting a plurality of input data symbol sub-streams with said approximation of said transmit weight matrix for distributing said plurality of input data symbol sub-streams along said plurality of spatial-subspace channels, said plurality of input data symbol sub-streams being derived from said input data symbol stream.

18. The method of claim 17, wherein said partial SVD algorithm comprises:

c) performing n1 iterations of an iterative SVD-algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, said interim weight matrix V_k′

being said interim version of said transmit weight matrix;

d) determining a number p of said plurality of spatial-subspace channels; and

,

e) truncating and zero-padding said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

19. The method of claim 18, wherein, said partial SVD algorithm further comprises:

f) performing n2 iterations of said iterative SVD algorithm on said truncated matrices U_k″

, Λ

_k″ and

V_k″

to obtain an approximate singular value matrix Λ

_k and said approximation of said transmit weight matrix V_k.

20. The method of claim 19, wherein n1 and n2 are constant integers.

21. The method of claim 20, wherein said constant integers are chosen from the range comprising integers 1 to 4.

22. The method of claim 18, wherein n1 is a variable integer bounded by a constant nmax1, and n1 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said interim matrix Λ

_k′

is less than a pre-specified fraction of a sum of squared magnitudes of diagonal components of said interim matrix Λ

_k′

.

23. The method of claim 19, wherein n2 is a variable integer bounded by a constant nmax2, and n2 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said truncated matrix Λ

_k″

is less than a pre-specified fraction of a sum of squared magnitudes of diagonal components of said truncated matrix Λ

_k″

.

24. The method of claim 19, wherein said method further comprises applying steps (a) to (f) of said partial SVD algorithm to said approximate singular value matrix Λ

_k.

25. A method for processing received spatial-subspace data on a plurality of spatial-subspace channels of a sub-carrier, wherein said method comprises:

a) calculating an approximation of a receive weight matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm which comprises truncating an interim version of said receive weight matrix to calculate said approximation of said receive weight matrix; and

,

b) weighting said received spatial-subspace data with said approximation of said receive weight matrix.

26. The method of claim 25, wherein said partial SVD algorithm comprises:

c) performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, a complex-conjugate transpose of said interim weight matrix U_k′

being said interim version of said receive weight matrix;

d) determining a number p of said plurality of spatial-subspace channels; and

,

e) truncating and zero-padding said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

27. The method of claim 26, wherein, said partial SVD algorithm further comprises:

f) performing n2 iterations of said iterative SVD algorithm on said truncated matrices U_k″

, Λ

_k″ and

V_k″

to obtain an approximate singular value matrix Λ

_k and said approximation of said receive weight matrix U_k*.

28. The method of claim 27, wherein n1 and n2 are constant integers.

29. The method of claim 28, wherein said constant integers are chosen from the range comprising integers 1 to 4.

30. The method of claim 26, wherein n1 is a variable integer bounded by a constant nmax1, and n1 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said interim matrix Λ

_k′

is less than a pre-specified fraction of a sum of squared magnitudes of diagonal components of said interim matrix Λ

_k′

.

31. The method of claim 27, wherein n2 is a variable integer bounded by a constant nmax2, and n2 is determined by performing said iterative SVD algorithm until a sum of squared magnitudes of off-diagonal components of said truncated matrix Λ

_k″

is less than a pre-specified fraction of a sum of squared magnitudes of diagonal components of said truncated matrix Λ

_k″

.

32. The method of claim 27, wherein said method further comprises applying steps (a) to (f) of said partial SVD algorithm to said approximate singular value matrix Λ

_k.

33. A communication system comprising a transmitter having a transmitter antenna array and a receiver having a receiver antenna array, wherein said communication system transmits an input data symbol stream over a plurality of spatial-subspace channels of a sub-carrier between said transmitter and said receiver, said input data symbol stream being partitioned into a plurality of super-frames, wherein said transmitter further comprises:

a) a subspace training unit for inserting subspace training symbols into a plurality of input data symbol sub-streams and producing a plurality of input data/training symbol sub-streams, said plurality of input data symbol sub-streams being derived from said input data symbol stream;

b) a transmitter SVD unit for providing a transmit weight matrix corresponding to said sub-carrier at a beginning of each super-frame in said plurality of super-frames; and

,

c) a transmitter weighting unit connected to said subspace training unit and said transmitter SVD unit for weighting said plurality of input data/training symbol sub-streams with said transmit weight matrix for distributing said plurality of input data/training symbol sub-streams along said plurality of spatial-subspace channels and providing transmit-weighted spatial-subspace data;

wherein, said receiver periodically processes said sub-space training symbols during each super-frame of said plurality of super-frames for estimating output data related to said input data symbol stream.

34. The communication system of claim 33, wherein said receiver further comprises:

d) a receiver SVD unit connected for providing a receive weight matrix corresponding to said sub-carrier and an initial estimate of a corresponding singular value matrix at a beginning of each super-frame of said plurality of super-frames;

e) a receiver weighting unit connected to said receiver SVD unit for receiving received spatial-subspace data related to said transmit-weighted spatial-subspace data and weighting said received spatial-subspace data with said receive weight matrix for providing receive-weighted spatial-subspace data, said receive-weighted spatial-subspace data containing received subspace training symbols;

f) a receiver link adaptation unit connected to said receiver weighting unit and said receiver SVD unit for receiving said initial estimate of said corresponding singular value matrix and processing said received subspace training symbols for periodically providing updated estimates of said corresponding singular value matrix during each super-frame of said plurality of super-frames; and

,

g) a data estimation unit connected to said receiver weighting unit and said receiver link adaptation unit for receiving said receive-weighted spatial-subspace data, said initial estimate of said corresponding singular value matrix and said updated estimates of said singular value matrix and providing said estimated output data during each super-frame of said plurality of super-frames.

35. The communication system of claim 33, wherein said subspace training unit inserts said subspace training symbols into only one of said plurality of input data symbol sub-streams at a time for tracking variations in a corresponding one of said plurality of spatial-subspace channels.

36. The communication system of claim 35, wherein said subspace training symbols are weighted by said transmit weight matrix and by said receive weight matrix thereby providing updated information on one of said plurality of spatial-subspace channels.

37. The communication system of claim 33, wherein said subspace training symbols have a low peak-to-average power.

38. The communication system of claim 33, wherein said transmitter further comprises a transmitter channel unit connected to said transmitter weighting unit for inserting channel training symbols into said transmit-weighted spatial-subspace data, wherein said channel training symbols and said subspace training symbols are derived from a substantially similar training sequence.

39. The communication system of claim 34, wherein said transmitter further comprises an encoder unit connected to said subspace training unit for processing said plurality of input data symbol sub-streams for providing at least one of:

1) at least one uncoded data symbol sub-stream for allocation on a corresponding at least one uncoded spatial-subspace channel and

     2) at least one pair of coded data symbol sub-streams for allocation on a corresponding at least one pair of coded spatial-subspace channels.

40. The communication system of claim 39, wherein said plurality of spatial-subspace channels comprise only uncoded spatial-subspace channels and said data estimation unit is adapted to perform a successive interference cancellation method for providing said estimated output data by detecting receive-weighted data symbols on one of said uncoded spatial-subspace channels, subtracting the detected receive-weighted data symbols from the receive-weighted spatial-subspace data and processing remaining uncoded spatial-subspace channels in a similar fashion wherein the uncoded spatial-subspace channels are processed in order of decreasing signal to noise plus interference ratio.

41. The communication system of claim 39, wherein said plurality of spatial-subspace channels comprise coded spatial-subspace channels and uncoded spatial-subspace channels and said data estimation unit is adapted to perform an iterative decoding/detection method on the receive-weighted spatial-subspace data for providing said estimated output data, with one iteration of said decoding/detection method comprising:

1) decoding and detecting receive-weighted coded data symbols transmitted on said coded spatial-subspace channels and replacing said receive-weighted coded data symbols in said receive-weighted spatial-subspace data with detected coded data symbols; and

,

     2) estimating and detecting receive-weighted uncoded data symbols transmitted on said uncoded spatial-subspace channels.

42. The communication system of claim 33, wherein said transmitter SVD unit calculates an approximation of said transmit weight matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm in which said transmitter SVD unit truncates an interim version of said transmit weight matrix to calculate said approximation of said transmit weight matrix.

43. The communication system of claim 42, wherein said transmitter SVD unit performs said partial SVD algorithm by performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, said interim weight matrix V_k′

being said interim version of said transmit weight matrix, wherein said transmitter SVD unit further determines a number p of said plurality of spatial-subspace channels, and truncates and zero-pads said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

44. The communication system of claim 43, wherein said transmitter SVD unit further performs n2 iterations of said iterative SVD algorithm on said truncated matrices U_k″

, Λ

_k″ and

V_k″

to obtain an approximate singular value matrix Λ

_k and said approximation of said transmit weight matrix V_k.

45. The communication system of claim 44, wherein said transmitter SVD unit further applies the partial SVD algorithm to said approximate singular value matrix Λ

_k.

46. The communication system of claim 34, wherein said receiver SVD unit calculates an approximation of said receive weight matrix and an approximation of said initial estimate of said corresponding singular value matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm in which said receiver SVD unit truncates interim versions of said receive weight matrix and said initial estimate of said corresponding singular value matrix.

47. The communication system of claim 46, wherein said receiver SVD unit performs said partial SVD algorithm by performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, a complex conjugate transpose of said interim weight matrix U_k′

being said interim version of said receive weight matrix and said interim singular value matrix Λ

_k′

being said interim version of said initial estimate of said corresponding singular value matrix, wherein said receiver SVD unit further determines a number p of said plurality of spatial-subspace channels and truncates and zero-pads said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

48. The communication system of claim 47, wherein said receiver SVD unit further performs n2 iterations of said iterative SVD algorithm on said truncated matrices Λ

_k″

, U_k″ and

V_k″

to obtain said approximation of said initial estimate of said corresponding singular value matrix Λ

_k and said approximation of said receive weight matrix U_k*.

49. The communication system of claim 48, wherein said receiver SVD unit further applies the partial SVD algorithm to said approximation of said initial estimate of said corresponding singular value matrix Λ

_k.

50. A data structure for transmitting data over a plurality of spatial-subspace channels of a sub-carrier between a transmitter and a receiver, said data structure comprising a plurality of super-frames, each of said plurality of super-frames comprising a channel training block followed by a plurality of frames, each of said plurality of frames comprising:

a) a slot for at least a portion of a sub-space training sequence;

b) a slot for at least a portion of a synchronization sequence; and

,

c) a plurality of slots for data sequences,

wherein, said sub-space training sequence allows said receiver to track variations in one of said plurality of spatial-subspace channels.

51. The data structure of claim 50, wherein said data structure further comprises a slot for at least a portion of channel/transmission information sequence for providing channel related data for a channel corresponding to said sub-carrier.

52. The data structure of claim 50, wherein said channel training block comprises a plurality of channel training sequences, said channel training sequences and said sub-space training sequences being derived from a substantially similar training sequence.

53. The data structure of claim 50, wherein one slot alternately contains said sub-space training sequence and said synchronization sequence.

54. A method for transmitting an input data symbol stream over a plurality of spatial-subspace channels of a sub-carrier between a transmitter and a receiver, said input data stream being partitioned into a plurality of super-frames, wherein at the transmitter said method comprises:

a) inserting subspace training symbols into a plurality of input data symbol sub-streams and producing a plurality of input data/training symbol sub-streams, said plurality of input data symbol sub-streams being derived from said input data symbol stream;

b) providing a transmit weight matrix corresponding to said sub-carrier at a beginning of each super-frame in said plurality of super-frames; and

,

c) weighting said plurality of input data/training symbol sub-streams with said transmit weight matrix for distributing said plurality of input data/training symbol sub-streams along said plurality of spatial-subspace channels and providing transmit-weighted spatial-subspace data;

wherein, said receiver periodically processes said sub-space training symbols during each super-frame of said plurality of super-frames for estimating output data related to said input data symbol stream.

55. The method of claim 54, wherein at the receiver, said method further comprises:

d) providing a receive weight matrix corresponding to said sub-carrier and an initial estimate of a corresponding singular value matrix at a beginning of each super-frame of said plurality of super-frames;

e) receiving received spatial-subspace data related to said transmit-weighted spatial-subspace data and weighting said received spatial-subspace data with said receive weight matrix for providing receive-weighted spatial-subspace data, said receive-weighted spatial-subspace data containing received subspace training symbols;

f) periodically providing updated estimates of said corresponding singular value matrix by processing said received subspace training symbols during each super-frame of said plurality of super-frames; and

,

g) providing said estimated output data by processing said receive-weighted spatial-subspace data using said initial estimate of said corresponding singular value matrix and said updated estimates of said singular value matrix during each super-frame of said plurality of super-frames.

56. The method of claim 54, wherein step (a) comprises inserting said subspace training symbols into only one of said plurality of input data symbol sub-streams at a time for tracking variations in a corresponding one of said plurality of spatial-subspace channels.

57. The method of claim 54, wherein step (c) further comprises inserting channel training symbols into said transmit-weighted spatial-subspace data, wherein said channel training symbols and said subspace training symbols are derived from substantially similar training sequences.

58. The method of claim 55, wherein step (a) further comprises processing said plurality of input data symbol sub-streams for providing at least one of:

1) at least one uncoded data symbol sub-stream for allocation on a corresponding at least one uncoded spatial-subspace channel and

     2) at least one pair of coded data symbol sub-streams for allocation on a corresponding at least one pair of coded spatial-subspace channels.

59. The method of claim 58, wherein said plurality of spatial-subspace channels comprise only uncoded spatial-subspace channels and step (g) comprises performing a successive interference cancellation method for providing said estimated output data by detecting receive-weighted data symbols on one of said uncoded spatial-subspace channels, subtracting the detected receive-weighted data symbols from the receive-weighted spatial-subspace data and processing remaining uncoded spatial-subspace channels in a similar fashion wherein the uncoded spatial-subspace channels are processed in order of decreasing signal to noise plus interference ratio.

60. The method of claim 58, wherein said plurality of spatial-subspace channels comprise coded spatial-subspace channels and uncoded spatial-subspace channels and step (g) comprises performing an iterative decoding/detection method on the receive-weighted spatial-subspace data for providing said estimated output data, with one iteration of said decoding/detection method comprising:

1) decoding and detecting of receive-weighted coded data symbols transmitted on said coded spatial-subspace channels and replacing said receive-weighted coded data symbols in said receive-weighted spatial-subspace data with detected coded data symbols; and

,

     2) estimating and detecting receive-weighted uncoded data symbols transmitted on said uncoded spatial-subspace channels.

61. The method of claim 54, wherein step (b) comprises calculating an approximation of said transmit weight matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm which comprises truncating an interim version of said transmit weight matrix.

62. The method of claim 61, wherein said partial SVD algorithm comprises performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, said interim weight matrix V_k′

being said interim version of said transmit weight matrix, wherein said partial SVD algorithm further comprises determining a number p of said plurality of spatial-subspace channels, and truncating and zero-padding said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

63. The method of claim 62, wherein said partial SVD algorithm further comprises performing n2 iterations of said iterative SVD algorithm on said truncated matrices U_k″

, Λ

_k″ and

V_k″

to obtain an approximate singular value matrix Λ

_k and said approximation of said transmit weight matrix V_k.

64. The method of claim 55, wherein step (d) comprises calculating an approximation of said receive weight matrix and an approximation of said initial estimate of said corresponding singular value matrix from a channel matrix corresponding to said sub-carrier by applying a partial SVD algorithm which comprises truncating an interim version of said receive weight matrix and said initial estimate of said corresponding singular value matrix.

65. The method of claim 64, wherein said partial SVD algorithm comprises performing n1 iterations of an iterative SVD algorithm on said channel matrix to obtain an interim singular value matrix Λ

_k′ and

interim weight matrices U_k′ and

V_k′

, a complex conjugate transpose of said interim weight matrix U_k′

being said interim version of said receive weight matrix and said interim singular value matrix Λ

_k′

being said interim version of said initial estimate of said corresponding singular value matrix, wherein said partial SVD algorithm further comprises determining a number p of said plurality of spatial-subspace channels and truncating and zero-padding said interim matrices Λ

_k′

, U_k′ and

V_k′

to obtain truncated matrices Λ

_k″

, U_k″ and

V_k″

.

66. The method of claim 65, wherein said partial SVD algorithm further comprises performing n2 iterations of said iterative SVD algorithm on said truncated matrices Λ

_k″

, U_k″ and

V_k″

to obtain an approximation of said initial estimate of said corresponding singular value matrix Λ

_k and said approximation of said receive weight matrix U_k*.