Beam-former and combiner for a multiple-antenna system

• US 20040219899A1
• Filed: 04/30/2003
• Published: 11/04/2004
• Est. Priority Date: 04/30/2003
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Ho, Minnie, Sandhu, Sumeet
• Granted Patent: US 7,746,967 B2
• Time in Patent Office: Days
• US Class Current: 455/273
• CPC Class Codes: H04B 7/0842 Weighted combining
• CPC Class Codes: H04B 7/0851 using training sequences or error signal minimizing error signal H04B7/0854
• CPC Class Codes: H04B 7/0854 using error minimizing algorithms, e.g. minimum mean squared error [MMSE], "cross-correlation" or matrix inversion

Abstract

A compensating correction value for adjusting analog signals received from multiple antenna elements takes into account the effects of colored noise, co-channel interference, and inter-sample interference. The method of generating the compensating correction value for analog combining architectures considers the total channel impulse response over a block of time.

Claims

1. A method, comprising:

computing weights for analog signals from multiple antenna elements of a receiver that combine multi-path signals and inter-sample interference.

2. The method of claim 1 wherein computing weights includes:

using quadrature signals derived from the analog signals to form a first matrix H;

forming a second matrix H* by multiplying the first matrix H with a conjugate transpose of the first matrix H; and

performing a Singular Value Decomposition (SVD) algorithm on the matrix HH*.

3. The method of claim 1 wherein computing weights includes:

using quadrature signals derived from the analog signals to form a first matrix H;

averaging the first matrix H to generate a second matrix Ha; and

performing the Singular Value Decomposition (SVD) algorithm on the second matrix Ha.

4. The method of claim 1 wherein computing weights includes:

using quadrature signals derived from the analog signals to form a first matrix H;

forming a matrix N based on colored noise estimates;

forming a new matrix by multiplying the matrix N with a conjugate transpose N*;

computing the square root of NN*;

pre and post multiplying the first matrix H by an inverse of a square root of NN* to provide a resulting matrix; and

performing a Singular Value Decomposition (SVD) algorithm on the resulting matrix.

5. The method of claim 1 wherein computing weights includes:

using quadrature signals to estimate interference and noise in channels in the absence of the analog signals to form a matrix N; and

using an iterative algorithm on the matrix N.

6. The method of claim 1 further comprising:

controlling a gain of a Low Noise Amplifier in a receive chain with the computed weights.

7. The method of claim 1 further comprising:

controlling a phase of a phase shifter in a receive chain with the computed weights.

8. The method of claim 1 further comprising:

controlling a phase of a phase shifter and a gain of a power amplifier in a transmitter with the computed weights.

9. A method, comprising:

sending known symbols to a receiver having multiple antenna to derive a compensating correction value that may be applied to a receiver chain that corresponds to each antenna.

10. The method of claim 9 further including:

computing weights for beam-forming analog signals through the multiple antenna at the receiver that combines multi-path signals and inter-sample interference.

11. The method of claim 9 further including:

estimating a channel impulse response in a time-domain to provide the compensating correction value as an analog value.

12. The method of claim 9 further including:

controlling signal amplification provided by a Low Noise Amplifier and phase through a phase shifter in a receive chain with the compensating correction value.

13. The method of claim 9 further comprising:

controlling a phase of a phase shifter and a gain of a power amplifier in a transmitter with the compensating correction value.

14. A system comprising:

a Static Random Access Memory (SRAM);

a communication processor coupled via an external bus to the SRAM, where the communication processor receives modulated signals at multiple antenna and computes a compensating correction value for analog signals from multiple antenna elements that combine multi-path signals and inter-sample interference; and

amplifier circuits in a receiver to receive the modulated signals and further receiving the compensating correction value.

15. The system of claim 14, further including:

phase shifters in the receiver that are coupled to the amplifier circuits, wherein the phase shifters receive the compensating correction value.

16. The system of claim 14, further including:

power amplifiers in a transmitter to receive the compensating correction value, wherein the power amplifiers provide gain to signals transmitted from the multiple antenna elements.

17. A method of computing a compensating correction value for analog signals from multiple antenna elements, comprising:

using quadrature signals derived from the analog signals to estimate interference and noise in channels; and

using the quadrature signals to form a matrix H.

18. The method of claim 17, further comprising:

forming a matrix K based on colored noise estimates;

forming a new matrix KK* by multiplying the matrix K with a conjugate transpose K*; and

computing a square root of the new matrix KK*.

19. The method of claim 18, further comprising:

pre and post multiplying the matrix H by an inverse of the square root of KK* to provide a resulting matrix.

20. The method of claim 19, further comprising:

performing a Singular Value Decomposition (SVD) algorithm on the resulting matrix to generate the compensating correction value.

21. The method of claim 17, further comprising:

forming a matrix K based on colored noise estimates; and

generating the compensating correction value based on the Minimum Mean-Squared Error (MMSE) algorithm.

22. The method of claim 17, further comprising:

forming a matrix O based on interference and noise in the channels in the absence of any signal;

forming a new matrix OO* by multiplying the matrix O with a conjugate transpose O*; and

computing a square root of the new matrix OO*.

23. The method of claim 22, further comprising:

pre and post multiplying the matrix H by an inverse of the square root of OO* to provide a resulting matrix.

24. The method of claim 23, further comprising:

performing a Singular Value Decomposition (SVD) algorithm on the resulting matrix to generate the compensating correction value.

25. The method of claim 17, further comprising:

forming a matrix O based on interference and noise in the channels in the absence of any signal; and

computing the compensating correction value based on the Minimum Mean-Squared Error (MMSE) algorithm.

26. The method of claim 17, further comprising:

forming a matrix T based on user interference in a Spatial Division Multiple Access (SDMA) communications network;

forming a new matrix TT* by multiplying the matrix T with a conjugate transpose T*; and

computing a square root of the new matrix TT*.

27. The method of claim 26, further comprising:

pre and post multiplying the matrix H by an inverse of the square root of TT* to provide a resulting matrix.

28. The method of claim 27, further comprising:

performing a Singular Value Decomposition (SVD) algorithm on the resulting matrix to generate the compensating correction value.

29. The method of claim 17, further comprising:

forming a matrix T based on user interference in a Spatial Division Multiple Access (SDMA) communications network; and

computing the compensating correction value based on the Minimum Mean-Squared Error (MMSE) algorithm.