Channel-matrix reduction for precoding control

1. A method of controlling multi-antenna transmission, the method comprising:

determining a channel matrix representing characteristics of a multi-path channel between a transmitter device equipped with multiple transmitter antennas and a receiver device equipped with multiple receiver antennas, the channel matrix being organized in a first number of channel vectors, each associated with a different one of the multiple receiver antennas;

combining the channel vectors to a smaller second number of linear combinations of the channel vectors;

composing a reduced channel matrix from the linear combinations of the channel vectors;

determining a precoding matrix based on the reduced channel matrix; and

controlling multi-antenna transmission by the transmitter device based on the determined precoding matrix;

wherein said combining of the channel vectors and composing of the reduced channel matrix comprises:

determining a Gram matrix of the channel matrix;

determining eigenvectors and associated eigenvalues of the Gram matrix;

selecting the eigenvectors associated with the largest eigenvalues; and

transforming the channel matrix based on the selected eigenvectors.

2. The method according to claim 1 , wherein a number of the selected eigenvectors corresponds to the second number.

3. The method according to claim 1 , further comprising:

composing a transformation matrix from the selected eigenvectors;

by applying the transformation matrix to the channel matrix, transforming the channel matrix to the reduced channel matrix.

4. The method according to claim 1 , further comprising:

determining the channel matrix for each of multiple different frequencies;

determining the Gram matrix by averaging over states of the channel matrix as determined for each of the multiple different frequencies.

5. The method according to claim 4 , wherein the multi-antenna transmission is based on modulation of symbols onto multiple subcarriers and the multiple frequencies comprise at least some of the subcarriers.

6. The method according to claim 1 , wherein said determining of the precoding matrix is based on a minimum mean square error based algorithm or on a zero-forcing based algorithm.

7. The method according to claim 1 , comprising:

determining the channel matrix based on measurements on reference signals transmitted from the receiver device to the transmitter device.

8. The method according to claim 1 , comprising:

determining the channel matrix based on measurements on reference signals transmitted from the transmitter device to the receiver device.

9. The method according to claim 1 , wherein the transmitter device is an access node of a wireless communication network.

10. The method according to claim 1 , wherein the receiver device is a wireless communication device connected to a wireless communication network.

11. A device, the device, comprising

at least one processor, and

a memory storing program code, which when executed by the at least one processor, causes the device to:

determine a channel matrix representing characteristics of a multi-path channel between a transmitter device equipped with multiple transmitter antennas and a receiver device equipped with multiple receiver antennas, the channel matrix being organized in a first number of channel vectors, each associated with a different one of the multiple receiver antennas;

combine the channel vectors to a smaller second number of linear combinations of the channel vectors;

compose a reduced channel matrix from the linear combinations of the channel vectors;

determine a precoding matrix based on the reduced channel matrix; and

control multi-antenna transmission by the transmitter device based on the determined precoding matrix;

wherein combining the channel vectors and composing the reduced channel matrix comprises:

determining a Gram matrix of the channel matrix;

determining eigenvectors and associated eigenvalues of the Gram matrix;

selecting the eigenvectors associated with the largest eigenvalues; and

transforming the channel matrix based on the selected eigenvectors.

12. The device according to claim 11 , wherein a number of the selected eigenvectors corresponds to the second number.

13. The device according to claim 11 , wherein the device is further configured to combine the channel vectors and compose the reduced channel matrix by:

composing a transformation matrix from the selected eigenvectors;

by applying the transformation matrix to the channel matrix, transforming the channel matrix to the reduced channel matrix.

14. The device according to claim 11 , wherein the device is further configured to:

determine the channel matrix for each of multiple different frequencies;

determine the Gram matrix by averaging over states of the channel matrix as determined for each of the multiple different frequencies.

15. The device according to claim 14 , wherein the multi-antenna transmission is based on modulation of symbols onto multiple subcarriers and the multiple frequencies comprise at least some of the subcarriers.

16. The device according to claim 11 , wherein the device is configured to determine the precoding matrix is based on a minimum mean square error based algorithm or on a zero-forcing based algorithm.

17. The device according to claim 11 , wherein the device is configured to determine the channel matrix based on measurements on reference signals transmitted from the receiver device to the transmitter device.

18. The device according to claim 11 , wherein the device is configured to determine the channel matrix based on measurements on reference signals transmitted from the transmitter device to the receiver device.

19. The device according to claim 11 , wherein the transmitter device is an access node of a wireless communication network.

20. The device according to claim 11 , wherein the receiver device is a wireless communication device connected to a wireless communication network.