Multi input multi output wireless communication method and apparatus providing extended range and extended rate across imperfectly estimated channels

• US 7,260,153 B2
• Filed: 04/28/2003
• Issued: 08/21/2007
• Est. Priority Date: 09/09/2002
• Status: Active Grant

• Current Assignee: Fleet Connect Solutions LLC (Empire IP LLC)
• Original Assignee: Mimopro Limited
• Inventors: Nissani (Nissensohn), Daniel Nathan
• Primary Examiner(s): Fan; Chieh M.
• Assistant Examiner(s): Lee; Siu M.
• Application Number:
  US10/423,447
• Publication Number:
  US 20040047426A1
• Time in Patent Office: 1,576 Days
• Field of Search: 375/267, 375/347, 375/285, 375/346
• US Class Current: 375/259
• CPC Class Codes: H04L 25/0204 of multiple channels
• CPC Class Codes: H04L 25/0242 using matrix methods

Abstract

A Multi Input Multi Output (MIMO) Wireless Communication System method and apparatus are proposed whereby in a 2-way wireless communication system with scattering, random and imperfectly estimated propagation channels the ubiquitous and inherent MIMO cross-talk interference problem is solved so that robust and predictable Extended Communication Range and Extended Data Rate are achieved.

Claims

1. A method for evaluating a channel of a multiple-input multiple-output (MIMO) wireless communication system, wherein said communication system comprises at least two communication devices having a plurality of radiating elements for the parallel transmission of data sub-streams, comprising:

defining a channel matrix metric, said channel matrix metric comprising a respective predefined function of channel matrix singular values for each of said data sub-streams, such that each of said predefined functions provides a measure of cross-talk signal to noise ratio (SNR) for said respective sub-stream;

obtaining an estimated channel matrix;

performing a singular value decomposition (SVD) of said estimated channel matrix to obtain estimated channel singular values, said singular value decomposition comprising a left-hand unitary weighting matrix, a diagonal matrix of said estimated channel singular values, and a right-hand unitary weighting matrix; and

calculating a respective crosstalk measure for each of said sub-streams from said channel matrix metric and said estimated channel singular values.

2. A method according to claim 1, wherein each of said functions maps said channel singular values onto an expected value of said cross-talk SNR of said respective sub-stream.

3. A method according to claim 1, wherein each of said functions maps said channel singular values onto a median value of said cross-talk SNR of said respective sub-stream.

4. A method according to claim 1, wherein each of said functions maps said channel singular values onto a value such that a probability of a channel cross-talk SNR of said respective sub-stream being greater than said value equals a pre-determined threshold.

5. A method according to claim 4, further comprising approximating said predefined functions as iso-valued polynomial surface functions of channel singular values, wherein each of said iso-valued surfaces divides the linear space defined by said singular values into a respective two partitions.

6. A method according to claim 1, further comprising determining the necessity of a pre-equalizer matrix from said calculated crosstalk measures.

7. A method according to claim 5, further comprising deriving a pre-equalizer matrix so as to yield a modified channel matrix having singular values within an intersection of said partitions, wherein said modified channel matrix comprises said estimated channel matrix pre-multiplied by said pre-equalizer matrix.

8. A method according to claim 7, wherein said intersection of partitions comprises a set of singular values such that the value of each of said iso-valued polynomial surface functions, evaluated at any point within said set of singular values, is no less than zero.

9. A method according to claim 7, wherein said deriving comprises:

providing a desired set of singular values of said modified channel matrix, wherein said set of singular values lies within said intersection of partitions;

forming a diagonal matrix of said desired singular values; and

performing a matrix multiplication of said right hand unitary weighting matrix, an inverse of said diagonal matrix of said estimated channel singular values, and said diagonal matrix of said desired singular values.

10. A method according to claim 1, further comprising deriving a pre-equalizer matrix from said matrix metric so as to yield a modified channel matrix having singular values Which maximize a weighted sum of net total SNR gains of said sub-streams, wherein said modified channel matrix comprises said estimated channel matrix pre-multiplied by said pre-equalizer matrix.

11. A method according to claim 1, further comprising deriving a pre-equalizer matrix from said matrix metric so as to yield a modified channel matrix with improved crosstalk measures, wherein said modified channel matrix comprises said estimated channel matrix pre-multiplied by said pre-equalizer matrix.

12. A method according to claim 11, further comprising calculating a diagonal power allocation matrix for allocating radiated transmission power such that a total prescribed maximum transmission power is preserved.

13. A method according to claim 12, wherein the diagonal elements of said diagonal power allocation matrix are equal.

14. A method according to claim 12, wherein the diagonal elements of said diagonal power allocation matrix are proportional to respective channel matrix singular values.

15. A method according to claim 12, wherein said pre-equalizer matrix and said power allocation matrix are calculated so as to obtain cross-talk SNRs not less than specified thresholds, and so as to minimize a resultant SNR loss from the application of said power allocation matrix.

16. A method according to claim 15, further comprising setting said pre-equalizer matrix to an identity matrix if an SNR loss resultant from said pre-equalizer matrix exceeds a specified threshold.

17. A method according to claim 10, further comprising setting said pre-equalizer matrix to an identity matrix if a net total SNR gain resultant from said pre-equalizer matrix is smaller than unity.

18. A method according to claim 11, further comprising defining a post-equalizer matrix by performing the steps of:

transmitting a training sequence over a wireless communication channel, said training sequence comprising a predefined data sequence multiplied by said pre-equalizer matrix prior to transmission;

receiving said transmitted data sequence;

multiplying said received data sequence by a complex transpose of said left-hand unitary weighting matrix to form a weighted received data signal;

for each data sub-stream, calculating a ratio of a transmitted data signal and a weighted received data signal; and

forming a diagonal matrix of said calculated ratios.

19. A method according to claim 1, wherein processing of said MIMO communication is performed in the time-domain.

20. A method according to claim 1, wherein processing of said MIMO communication is performed in the frequency-domain.

21. A method for communicating digital data over a MIMO wireless communication system, wherein said communication system comprises at least two communication devices having a plurality of radiating elements for the parallel transmission of data sub-streams, comprising:

obtaining an estimated channel matrix;

multiplying said data sub-streams by a unitary transmission weighting matrix to form a weighted data stream, wherein said transmission weighting matrix comprises a conjugate transpose of a right-hand unitary matrix of a singular value decomposition of an estimated channel matrix, and wherein said singular value decomposition comprises a left-hand unitary weighting matrix, a diagonal matrix of said estimated channel singular values and a right-hand unitary weighting matrix;

multiplying said weighted data stream by a pre-equalizer matrix to form a pre-equalizer data stream, said pre-equalizer matrix being determined by the steps of;

defining a matrix metric, said channel matrix metric comprising a respective predefined function of channel matrix singular values, for each of said data sub-streams, such that each of said predefined functions provides a measure of cross-talk signal to noise ratio (SNR) for said respective sub-stream;

calculating a respective crosstalk measure for each of said sub-streams from said channel matrix metric and said estimated channel singular values; and

calculating said pre-equalizer matrix so as to yield a modified channel matrix having improved crosstalk measures, wherein said modified channel matrix comprises said estimated channel matrix pre-multiplied by said pre-equalizer matrix;

and

transmitting said pre-equalized data stream.

22. A method according to claim 21, further comprising pre-multiplying said data sub-streams by a diagonal power allocation matrix, such that a total prescribed maximum transmission power is preserved.

23. A method according to claim 21, reducing a number of transmitted data sub-streams if an SNR loss resultant from said pre-equalizer matrix exceeds a specified threshold.

24. A method according to claim 21, reducing a number of transmitted data sub-streams if a net total SNR gain resultant from said pre-equalizer matrix is smaller than unity.

25. A method according to claim 21, further comprising approximating said predefined functions as iso-valued polynomial surface functions of channel singular values, wherein each of said iso-valued surfaces divides the linear space defined by said singular values into a respective two partitions, and wherein said modified channel matrix is calculated to have singular values within an intersection of said partitions.

26. A method according to claim 21, wherein said modified channel matrix is calculated to have singular values which maximize a weighted sum of net total SNR gains of said sub-streams.

27. A method according to claim 21, further comprising:

receiving said transmitted data;

multiplying received data sub-streams by a reception weighting matrix to form a weighted received data stream, wherein said reception weighting matrix comprises a complex conjugate transpose of said left-hand unitary weighting matrix of said estimated channel matrix; and

multiplying said weighted received data sub-streams by a post-equalizer matrix, said post-equalizer matrix being determined by the steps of;

transmitting a training sequence over a wireless communication channel, said training sequence comprising a predefined data sequence multiplied by said pre-equalizer matrix and said right-hand unitary matrix prior to transmission;

receiving said transmitted data signal;

multiplying said received data sequence by said reception weighting matrix to form a weighted received data signal;

for each data sub-stream, calculating a ratio of a transmitted data signal and a weighted received data signal; and

forming a diagonal matrix of said calculated ratios.

28. A transmission apparatus for evaluating a channel of a MIMO wireless communication system, wherein said communication system comprises at least two communication devices having a plurality of radiating elements for the parallel transmission of data sub-streams, comprising:

an SVD performer, configured for performing an SVD of an estimated channel matrix to obtain estimated channel singular values, said singular value decomposition comprising a left-hand unitary weighting matrix, a diagonal matrix of said estimated channel singular values and a right-hand unitary weighting matrix; and

a measure calculator associated with said SVD performer, configured for calculating a respective crosstalk measure for each of said sub-streams from said estimated channel singular values and a channel matrix metric, said channel matrix metric comprising a respective predefined function of channel matrix singular values for each of said data sub-streams, such that each of said predefined functions provides a measure of cross-talk signal to noise ratio (SNR) for said respective sub-stream.

29. An apparatus according to claim 28, wherein each of said functions maps said channel singular values onto an expected value of said cross-talk SNR of said respective sub-stream.

30. An apparatus according to claim 28, wherein each of said functions maps said channel singular values onto a median value of said cross-talk SNR of said respective sub-stream.

31. An apparatus according to claim 28, wherein each of said functions maps said channel singular values onto a value such that a probability of a channel cross-talk SNR of said respective sub-stream being greater than said value equals a pre-determined threshold.

32. An apparatus according to claim 28, wherein said predefined functions are approximated by iso-valued polynomial surface functions of channel singular values.

33. An apparatus according to claim 28, further comprising a pre-equalizer matrix deriver configured for deriving a pre-equalizer matrix so as to yield a modified channel matrix having improved crosstalk measures, wherein said modified channel matrix comprises said estimated channel matrix pre-multiplied by said pre-equalizer matrix.

34. An apparatus according to claim 33, wherein said pre-equalizer matrix is calculated so as to yield a modified channel matrix having singular values which maximize a weighted sum of net total SNR gains of said sub-streams.

35. An apparatus according to claim 33, wherein said pre-equalizer matrix is calculated so as to yield a modified channel matrix having singular values within an intersection of partitions of the linear space defined by said singular values, and wherein said partitions are created by iso-valued polynomial surface functions of channel singular values which approximate said predefined functions.

36. An apparatus according to claim 33, wherein said pre-equalizer matrix deriver is configured to derive said pre-equalizer matrix from a provided set of desired singular values of said modified channel matrix by performing a matrix multiplication of said right-hand unitary weighting matrix, an inverse of said diagonal matrix of said estimated channel singular values, and a diagonal matrix of said desired singular values.

37. An apparatus according to claim 33, further comprising a power allocation matrix deriver configured for calculating a diagonal power allocation matrix for allocating radiated transmission power such that a total prescribed maximum transmission power is preserved.

38. An apparatus according to claim 37, wherein said pre-equalizer matrix and said power allocation matrix are calculated so as to obtain cross-talk SNRs not less than specified thresholds, and so as to minimize a resultant SNR loss from the application of said pre-equalizer matrix and said power allocation matrix.

39. A transmitter for a MIMO wireless communication system comprising the apparatus of claim 33, and further comprising:

a transmission weighting matrix comprising said right-hand unitary weighting matrix, configured for multiplying a transmit data vector to form weighted data sub-streams; and

said pre-equalizer matrix being configured for multiplying said weighted data sub-streams to form a pre-equalized data stream for transmission.

40. The transmitter of claim 39, further comprising a diagonal power allocation matrix configured for allocating radiated transmission power between said sub-streams such that a total prescribed maximum transmission power is preserved.

41. A reception apparatus for a MIMO wireless communication system, wherein said communication system comprises at least two communication devices having a plurality of radiating elements for the parallel transmission of data sub-streams, comprising:

an SVD performer, configured for performing an SVD of an estimated channel matrix to obtain estimated channel singular values, said singular value decomposition comprising a left-hand unitary weighting matrix, a diagonal matrix of said estimated channel singular values and a right-hand unitary weighting matrix;

a post-equalizer matrix deriver configured for defining a post-equalizer matrix by receiving a training sequence transmitted over a wireless communication channel, multiplying said received data sequence by a complex transpose of said left-hand unitary weighting matrix to form a weighted received data signal, calculating, respectively for each data sub-stream, a ratio of a transmitted data signal and a received data signal, and forming a diagonal matrix of said calculated ratios;

a reception weighting matrix comprising a complex conjugate transpose of said right-hand unitary weighting matrix, configured for multiplying received data sub-streams to form weighted received data sub-streams; and

said post-equalizer matrix associated with said post-equalizer deriver and said reception weighting matrix, configured for multiplying weighted received data sub-streams to form a received data vector.