System and method for performing MLD preprocessing in a MIMO decoder

1. A method for performing Maximum Likelihood Detector (MLD) preprocessing in a Multiple-Input Multiple-Output (MIMO) communication system, the method comprising, using a processor:

obtaining a received signal, Y, a corresponding channel matrix, H and a vector of noise samples, n;

calculating a whitening filter L −H based on the vector of noise samples n;

whitening the channel matrix H using the whitening filter L −H to obtain a whitened channel matrix {tilde over (H)};

performing a coordinate rotation digital computer (Cordic) based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitening filter L −H to obtain triangular matrix R of the QR decomposition of the whitened channel matrix {tilde over (H)} and Q H L −H ; and

multiplying the received signal Y by Q H L −M to obtain Y =Q H L −H Y.

2. The method of claim 1 , wherein performing the Cordic based QR decomposition comprises performing the Cordic based QR decomposition to a widened matrix, the widened matrix being a combination of the whitened channel matrix {tilde over (H)} and the whitening filter L −H .

3. The method of claim 1 , wherein the MLD preprocessing is performed by a MIMO decoder.

4. The method of claim 1 , comprising decoding the received signal F by using the triangular matrix R and Y to determine a corresponding transmit signal s.

5. The method of claim 1 , wherein the whitening filter L −H is calculated by:

calculating a correlation matrix R nn of the vector of noise samples n; and

performing a Cholesky decomposition of the noise correlation matrix R nn to obtain the whitening filter L −H .

6. A method for performing Maximum Likelihood Detector (MLD) preprocessing in a Multiple-Input Multiple-Output (MIMO) communication system, the method comprising, using a processor:

obtaining a received signal Y a corresponding channel matrix H and a vector of noise samples n;

calculating a whitening filter L −M based on the vector of noise samples n;

whitening a channel matrix H using the whitening filter L −H to obtain a whitened channel matrix {tilde over (H)};

whitening the received signal Y using the whitening filter L −H to obtain a whitened received signal {tilde over (Y)}; and

performing a coordinate rotation digital computer (Cordic) based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitened received signal {tilde over (Y)} to obtain triangular matrix R of the QR decomposition of the whitened channel matrix {tilde over (H)} and Y =Q H L −H Y.

7. The method of claim 6 , wherein performing the Cordic based QR decomposition comprises performing the Cordic based QR decomposition to a widened matrix, the widened matrix being a combination of the whitened channel matrix {tilde over (H)}, and the whitened vector of the receiver antennas {tilde over (Y)}.

8. The method of claim 6 , wherein the MLD preprocessing is performed by a MIMO decoder.

9. The method of claim 6 , comprising decoding the received signal Y by using the triangular matrix R and Y to determine a corresponding transmit signal s.

10. The method of claim 6 , wherein the whitening filter L −H is calculated by:

calculating a correlation matrix R nn of the vector of noise samples n; and

performing a Cholesky decomposition of the noise correlation matrix R nn to obtain the whitening filter L −H .

11. A method for performing Maximum Likelihood Detector (MLD) preprocessing in a Multiple-Input Multiple-Output (MIMO) communication system, the method comprising, using a processor:

obtaining a received signal Y a corresponding channel matrix H and a vector of noise samples n;

calculating a whitening filter L −M based on the vector of noise samples n;

whitening a channel matrix H using the whitening filter L −H to obtain a whitened channel matrix {tilde over (H)};

selecting, one of a first calculation or a second calculation, based on estimated complexity of the first calculation and the second calculation; and

performing preprocessing of the received signal Y using the selected calculation,

wherein the first calculation comprises:

whitening the received signal Y using the whitening filter L −H to obtain a whitened received signal {tilde over (Y)} and performing a coordinate rotation digital computer (Cordic) based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitened received signal {tilde over (Y)} to obtain triangular matrix R of the QR decomposition of the whitened channel matrix {tilde over (H)} and Y =Q H L −H Y,

and wherein the second calculation comprises:

performing a Cordic based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitening filter L −H to obtain triangular matrix R of the QR decomposition of the whitened channel matrix {tilde over (H)} and Q H L −H , and

multiplying the received signal Y by Q H L −H to obtain Y =Q H L −H Y.

12. The method of claim 11 , wherein performing the Cordic based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitened received signal {tilde over (Y)} comprises performing the Cordic based QR decomposition to a widened matrix, the widened matrix being a combination of the whitened channel matrix {tilde over (H)} and the whitened received signal {tilde over (Y)}.

13. The method of claim 11 , wherein performing the Cordic based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitening filter L −H comprises performing the Cordic based QR decomposition to a widened matrix, the widened matrix being a combination of the whitened channel matrix {tilde over (H)} and the whitening filter L −H .

14. The method of claim 11 , wherein the MLD preprocessing is performed by a MIMO decoder.

15. The method of claim 11 , comprising decoding the received signal Y by using the triangular matrix R and Y to determine a corresponding transmit signal s.

16. The method of claim 11 , wherein the whitening filter, L −H is calculated by:

calculating a correlation matrix R nn of the vector of noise samples n; and

performing a Cholesky decomposition of the noise correlation matrix R nn to obtain the whitening filter L −H .

17. The method of claim 11 , comprising:

estimating the complexity of the first calculation and the second calculation based on a time domain decimation D_T, a frequency domain decimation D_F, and a number of transmitter antennas M.

18. A Multiple-Input Multiple-Output (MIMO) decoder comprising:

a memory;

a processor configured to:

obtain a received signal, Y, a corresponding channel matrix, H and a vector of noise samples, u;

calculate a whitening filter L −H based on the vector of noise samples n;

whiten the channel matrix H using the whitening filter L −H to obtain a whitened channel matrix {tilde over (H)};

perform a coordinate rotation digital computer (Cordic) based QR decomposition to the whitened channel matrix {tilde over (H)} and the whitening filter L −H to obtain triangular matrix R of the QR decomposition of the whitened channel matrix {tilde over (H)} and Q H L −H ;

multiply the received signal Y by Q H L −H to obtain Y =Q H L −H Y.

19. The MIMO decoder of claim 18 , wherein the processor is configured to perform the Cordic based QR decomposition by performing the Cordic based QR decomposition to a widened matrix, the widened matrix being a combination of the whitened channel matrix {tilde over (H)} and the whitening filter L −H .

20. The MIMO decoder of claim 18 , wherein the processor is configured to decode the received signal F by using the triangular matrix R and Y to determine a corresponding transmit signal s.