System and method for interference cancellation in a wireless communication receiver

• US 7,088,978 B2
• Filed: 02/07/2003
• Issued: 08/08/2006
• Est. Priority Date: 02/07/2003
• Status: Active Grant

• Current Assignee: Ericsson, Inc. (Telefonaktiebolaget LM Ericsson)
• Original Assignee: Ericsson, Inc. (Telefonaktiebolaget LM Ericsson)
• Inventors: Ramesh, Rajaram, Hui, Dennis
• Primary Examiner(s): Urban, Edward
• Assistant Examiner(s): CHOW, CHARLES CHIANG
• Application Number:
  US10/360,032
• Publication Number:
  US 20040198296A1
• Time in Patent Office: 1,278 Days
• Field of Search: 455/296, 455/63.1, 455/273.1, 455/501, 455/283, 455/67.13, 455/67.14, 455/67.16, 455/65, 455/506, 455/504, 455/10, 455/323, 455/364, 455/303, 455/525, 455/425, 455/424, 455/561, 455/456.5, 455/456.6, 455/67.11, 455/550.1, 455/295, 455/277.2, 455/278, 455/1, 370/263, 370/208, 370/337, 370/347, 370/319, 370/231, 375/232, 375/284, 375/285, 375/283, 375/348, 375/332, 375/341, 375/346, 375/227, 375/316, 375/148, 375/233, 375/243, 375/340, 375/352, 375/345, 375/254, 768/322, 768/323, 329/318, 329/320, 329/349, 329/353, 364/724, 367/135, 367/901, 367/138, 367/87, 704/233, 702/190, 702/191, 702/199, 702/181
• US Class Current: 455/296
• CPC Class Codes: H04L 2025/03401 PSK
• CPC Class Codes: H04L 2025/03407 Continuous phase
• CPC Class Codes: H04L 25/023 with extension to other symbols
• CPC Class Codes: H04L 25/025 using least-mean-square [LMS] method

Abstract

A communication receiver improves its channel estimation performance, its equalization performance, or both, by modeling the non-Gaussian characteristics of a dominant source of interference in a received signal. That is, rather than using Gaussian-based modeling, the receiver generates signal disturbance estimates using a probabilistic model representative of the non-Gaussian interference in a received signal, such as adjacent channel or co-channel signal interference in mobile station operating within a wireless communication network. The receiver may use such non-Gaussian interference modeling responsive to detecting such interference in the received signal and may match its disturbance model to the detected interference. Further, where signal disturbance is predominantly Gaussian in nature, the receiver may switch to a Gaussian disturbance model.

Claims

1. A method of improving performance in a wireless communication receiver comprising the steps of:

receiving a known signal at the receiver that is impaired by signal disturbance that includes non-Gaussian interference;

performing channel estimation to obtain a channel estimate based on the known signal and modeling the probabilistic characteristics of the non-Gaussian interference, comprising the steps of;

a) modeling the signal disturbance as Gaussian and non-Gaussian disturbance components by generating a disturbance estimate for the signal disturbance as

     1) a first, non-Gaussian component, comprising generating one or more parameters to model a non-Gaussian probability distribution function associated with the non-Gaussian interference, wherein the one or more parameters are based on knowledge of a signal type associated with the non-Gaussian interference, and

     2) a second. Gaussian component; and

,

b) performing channel estimation based on the disturbance estimate; and

, detecting an unknown signal received at the receiver using the channel estimate.

2. The method of claim 1, wherein receiving a known signal at the receiver comprises receiving a training sequence within a burst signal comprising the training sequence and unknown data such that detecting the unknown signal comprises detecting the unknown data in the burst signal.

3. The method of claim 1, wherein receiving a known signal at the receiver comprises receiving a pilot signal, and wherein detecting an unknown signal received at the receiver comprises demodulating traffic or control data transmitted in association with the pilot signal.

4. The method of claim 1, wherein the non-Gaussian interference is an interfering GMSK signal, and wherein generating the one or more parameters to model a non-Gaussian probability distribution of the non-Gaussian interference comprises defining an amplitude parameter to model a constant envelope of the interfering GMSK signal, and a phase parameter to model a phase of the interfering GMSK signal.

5. The method of claim 1, wherein generating a second, Gaussian component of the signal disturbance comprising generating a variance or covariance estimate for the signal disturbance.

6. The method of claim 1, wherein performing channel estimation to obtain a channel estimate using the known signal based on modeling the probabilistic characteristics of the non-Gaussian interference comprises:

initializing a channel estimate;

initializing a disturbance estimate for the known signal as a first, non-Gaussian component and a second, Gaussian component; and

revising the channel estimate based on the disturbance estimate.

7. The method of claim 6, further comprising iteratively obtaining revised channel and disturbance estimates.

8. The method of claim 7, further comprising iteratively revising the channel and disturbance estimates until an iteration criterion is satisfied.

9. The method of claim 8, wherein detecting an unknown signal received at the receiver using the channel estimate comprises detecting the unknown signal using the iteratively revised channel estimate.

10. The method of claim 6, wherein initializing a channel estimate comprises initially determining the channel estimate from the known signal using a Least Squares (LS) estimation technique that does not model the non-Gaussian interference impairing the known signal.

11. The method of claim 6, wherein initializing a channel estimate comprises evaluating the known signal including impairments in conjunction with nominal values of the known signal.

12. The method of claim 11, wherein revising the channel estimate based on the disturbance estimate comprises determining a difference between the known signal including impairments and the first component of the disturbance estimate.

13. The method of claim 11, further comprising revising the disturbance estimate as a function of the revised channel estimate to obtain a revised disturbance estimate for use in a subsequent revision of the channel estimate.

14. The method of claim 1, further comprising determining whether signals received at the receiver are impaired by non-Gaussian interference and selectively performing channel estimation based on modeling the non-Gaussian probability characteristics of the signal disturbance.

15. A wireless communication receiver comprising:

a receiver front-end to obtain a received signal from a received transmit signal;

a disturbance estimator to generate a disturbance estimate that models the probability characteristics of non-Gaussian interference impairing the received signal, comprising a receive processing circuit configured to;

generate a disturbance estimate for the non-Gaussian interference based on an initial channel estimate and a probabilistic model of the non-Gaussian interference; and

generate a residual signal based on the received signal and the disturbance estimate; and

a channel estimator to generate a channel estimate for the received signal based on the disturbance estimate, wherein the channel estimator is configured to compute a revised channel estimate using the residual signal and y obtain an error signal characterizing a channel estimate error, and the disturbance estimator is configured to compute a variance or covariance estimate and thereby update the disturbance estimate for subsequent generation of the residual signal and revision of the channel estimate.

16. The receiver of claim 15, wherein the channel estimator comprises a processing circuit configured to determine the channel estimate based on evaluating a residual value determined from a known portion of the received signal and the disturbance estimate.

17. The receiver of claim 16, wherein the disturbance and channel estimator cooperate to iteratively generate revised channel and disturbance estimates until an iteration criterion is satisfied.

18. The receiver of claim 15, wherein the disturbance estimator comprises:

a first estimation circuit to generate a disturbance estimate of non-Gaussian disturbance for the received signal;

a summing circuit to generate a first residual signal based on the received signal and the disturbance estimate; and

a second estimator to generate a variance estimate for the disturbance estimate based on an error signal characterizing an error of the channel estimate.

19. The receiver of claim 18, wherein the channel estimator comprises a channel estimation circuit to generate the channel estimate from the residual signal, and to generate the error signal.

20. A wireless communication receiver comprising:

a receiver front-end to obtain a received signal from a received transmit signal, wherein the received signal is impaired by signal disturbance including non-Gaussian interference;

a disturbance estimator to generate a disturbance estimate for the received signal based on modeling the probability characteristics of the non-Gaussian interference, comprising;

a first estimation circuit to generate a first, non-Gaussian component of the disturbance estimate based on modeling the probability distribution function of the non-Gaussian interference; and

a second estimation circuit to generate a second, Gaussian component of the disturbance estimate; and

an equalizer to generate a sequence estimate for the received signal based on an equalizer metric that includes a correction term based on the disturbance estimate.

21. The receiver of claim 20, wherein the equalizer comprises a portion of a baseband processing circuit.

22. The receiver of claim 20, wherein the equalizer defines the equalizer metric as a distance metric term and the correction term such that the correction term accounts for the non-Gaussian Interference in the received signal.

23. The receiver of claim 20, wherein the disturbance estimator comprises a receiver processing circuit configured to generate the disturbance estimate using a probabilistic model of the non-Gaussian interference.

24. The receiver of claim 20, wherein the disturbance estimator generates the second component of the disturbance estimate as a variance estimate for the signal disturbance.