Method for minimizing cross-talk in adaptive transmission antennas

• US 5,471,647 A
• Filed: 04/14/1993
• Issued: 11/28/1995
• Est. Priority Date: 04/14/1993
• Status: Expired due to Term

• Current Assignee: Board of Trustees of the Leland Stanford Junior University (Stanford Management Co.)
• Original Assignee: Leland Stanford Junior University (Stanford University)
• Inventors: Gerlach, Derek, Paulraj, Arogyaswami J.
• Primary Examiner(s): Eisenzopf, Reinhard J.
• Assistant Examiner(s): FAILE, ANDREW I
• Application Number:
  US08/046,631
• Time in Patent Office: 958 Days
• Field of Search: 455/62, 455/63, 455/67.3, 455/69, 455/70, 455/101, 455/103, 375/58, 375/60
• US Class Current: 455/63.1
• CPC Class Codes: H01Q 3/2605 Array of radiating elements provided with a feedback control over the element weights, e.g. adaptive arrays
• CPC Class Codes: H04B 1/126 having multiple inputs, e.g. auxiliary antenna for receiving interfering signal aerials in general H01Q
• CPC Class Codes: H04B 7/0617 for beam forming
• CPC Class Codes: H04B 7/0626 Channel coefficients, e.g. channel state information [CSI]

Abstract

A method is disclosed for using feedback signals (32) to optimize the directional properties of information signals (19) with respect to their intended receivers (16), where information signals (19) are applied to the transmitting elements (14) of a transmitting antenna array (12). Feedback signals (32) generated at receivers (16) by probing signals (17) are monitored, and the weight matrix with which signals (17) are distributed among array elements (14) is adjusted according to feedback signals (32) to minimize cross-talk. Information signals (18) transmitted according to an optimized weight matrix will also have reduced cross-talk. A weight matrix may be optimized by perturbing an initial weight matrix, calculating an error function from the feedback signals (32) produced by each perturbed weight matrix, estimating the direction of an extremum from the error functions, and adjusting the initial weight matrix in that direction. Alternatively, a gain matrix for the transmitting array may be explicitly determined by sequentially exciting each transmitting element (14) and measuring the resultant feedback signals (32). These are then used to calculate the weight matrix that eliminates cross-talk.

Claims

1. A method for minimizing cross-talk among information signals in a wireless communication system comprising an array of transmitting elements and a collection of receivers, the method comprising the steps of:

transmitting a set of probing signals from the array according to a probing weight matrix;

feeding back from the receivers to the array in response to the probing signals a set of feedback signals measuring the cross-talk;

calculating from the feedback signals a new weight matrix that reduces the cross-talk measured by the feedback signals; and

transmitting a set of new information signals according to the new weight matrix.

2. The method of claim 1 wherein the probing weight matrix is a perturbation of a selected weight matrix, and the feedback signals include error functions.

3. The method of claim 2 wherein the calculating step comprises:

computing an estimated gradient of the error functions; and

obtaining the new weight matrix by adjusting the selected weight matrix according to the estimated gradient.

4. The method of claim 3 wherein the set of probing signals comprises the information signals.

5. The method of claim 4 wherein the feeding back and the calculating steps are repeated until cross-talk is substantially minimized.

6. The method of claim 4 wherein the probing signals applied to the array using a type of multiplexing chosen from the group consisting of time, frequency, and code.

7. The method of claim 1 wherein the calculating step comprises the substeps of:

determining a gain matrix from the feedback signals; and

computing from the gain matrix the new weight matrix which substantially eliminates cross-talk.

8. The method of claim 7 wherein the determining substep comprises determining the gain matrix from a least squares fit to a collection of signal matrices.

9. The method of claim 8 wherein the computing substep comprises computing the new weight matrix from the pseudoinverse of the gain matrix.

10. The method of claim 1 wherein the calculating step comprises calculating the new weight matrix by multiplying the probing weight matrix by the Hermitian conjugate of the inverse of a signal matrix.

11. The method of claim 1 wherein the probing weight matrix is chosen so that each of the probing signals successively excites only one of the transmitting elements of the array.

12. A method for using feedback and spatial multiplexing to reduce cross-talk between information signals transmitted from an array of transmitting elements to a collection of receivers, the method comprising the steps of:

choosing a weight matrix W which characterizes the complex amplitudes with which the information signals are applied to the array of transmitting elements;

transmitting the information signals from the array to the receivers according to the weight matrix W;

feeding back from the receivers to the array a signal matrix C whose off-diagonal elements measure the cross-talk between the information signals;

calculating a new weight matrix W'"'"' from the weight matrix W and the signal matrix C such that a formula equivalent to W'"'"'=W(C^-1)^H is substantially satisfied; and

transmitting new information signals from the array according to the new weight matrix W'"'"'.

13. The method of claim 12 wherein the calculating step comprises the substeps of:

calculating a gain matrix A from the weight matrix W and the signal matrix C using a formula equivalent to W^H A=C;

calculating the new weight matrix W'"'"' from the gain matrix A using a formula equivalent to W'"'"'^H A=I or, equivalently, W'"'"'=A(A^H A)^-1.

14. The method of claim 12 wherein the weight matrix W is chosen so that all its elements are zero except for a diagonal collection of elements, each of which is equal to one.

15. The method of claim 12 wherein the calculating step comprises the substeps of:

calculating a gain matrix A from a least squares fit to a collection of signal matrices accumulated over time;

calculating the new weight matrix W'"'"' from the gain matrix A using a formula equivalent to W'"'"'^H A=I or, equivalently, W'"'"'=A(A^H A)^-1.

16. The method of claim 12 wherein the transmitting of the information signals comprises applying each of the information signals to the array according a weight vector corresponding to a column of the weight matrix W.

17. The method of claim 16 wherein the information signals are probing signals applied to the array using a type of multiplexing chosen from the group consisting of time, frequency, and code.

18. The method of claim 12 wherein the choosing of the weight matrix comprises perturbing a selected weight matrix;

and wherein the calculating step comprises using the signal matrix to determine the direction in which the weight matrix W should be adjusted to reduce cross-talk.