Device and method for supressing periodic interference signals

• US 20040242157A1
• Filed: 03/29/2004
• Published: 12/02/2004
• Est. Priority Date: 09/28/2001
• Status: Abandoned Application

• Current Assignee: Siemens AG
• Original Assignee: Siemens AG
• Inventors: Klinke, Stefano Ambrosius
• Application Number:
  US10/491,054
• Publication Number:
  US 20040242157A1
• Time in Patent Office: Days
• US Class Current: 455/63.1
• CPC Class Codes: G10L 2021/02085 Periodic noise
• CPC Class Codes: H04B 1/1036 with automatic suppression of narrow band noise or interference, e.g. by using tuneable notch filters H04B1/123 takes precedence; filter circuits H03H
• CPC Class Codes: H04B 15/005 Reducing noise, e.g. humm, from the supply

Abstract

The invention relates to a device and method for suppressing periodic interference signals, comprising a signal analyser (1), for the output of an error signal (d(k)) and corresponding coefficients (a_i) on the basis of a distorted wanted signal (x(k)), whereby a suppression unit (2) generates a suppressing error signal (d(k)), based on the error signal (d′(k)) and a signal synthesiser (3) regenerates a recovered wanted signal (x*(k)), based on the suppressing error signal (d′(k)) and the coefficients a_i.

Claims

1. A device for suppressing periodic interference signals, having

a signal analyzer (1) for outputting a fault signal (d(k)) and associated coefficients (a_i) on the basis of a useful signal (x(k)) which has been subjected to interference;

an interference suppressor unit (2) for generating a fault signal (d′

(k)) which has been subjected to interference suppression and has reduced periodic interference signals on the basis of the fault signal (d(k)); and

a signal synthesizer (3) for recovering a useful signal (x*(k)) which has been subjected to interference suppression, on the basis of the fault signal (d′

(k)) which has been subjected to interference suppression, and the coefficients (a_i).

2. The device as claimed in patent claim 1, characterized in that the signal analyzer (1) has an FIR filter and/or an IIR filter for outputting a predictive fault signal (d(k)) and associated predictor coefficients (a_i) on the basis of a voice signal (x(k)), and the signal synthesizer has an IIR filter and/or FIR filter for recovering the useful signal (x*(k)) which has been subjected to interference suppression, on the basis of a predictive fault signal ((d′

(k)) which has been subjected to interference suppression and associated predictor coefficients (a_i).

3. The device as claimed in patent claim 1 or 2, characterized in that the signal analyzer (1) has a linear predictor for carrying out a linear prediction.

4. The device as claimed in patent claim 3, characterized in that, the linear predictor (1) carries out a short-term prediction in a time domain of 20 to 400 milliseconds.

5. The device as claimed in one of patent claims 1 to 4, characterized in that the signal analyzer (1) determines the coefficients (a_i) by means of a Levinson-Durbin algorithm.

6. The device as claimed in one of patent claims 1 to 5, characterized by a high-pass filter (4) for filtering the useful signal (x(k)) which has been subjected to interference and for improving the coefficient calculation in the signal analyzer (1).

7. The device as claimed in patent claim 6, characterized in that the high-pass filter (4) has a preemphasis filter.

8. The device as claimed in one of patent claims 6 or 7, characterized by a low-pass filter (5) for filtering the useful signal (x*(k)) which has been subjected to interference suppression, and for compensating the high-pass filter (4).

9. The device as claimed in patent claim 8, characterized in that the low-pass filter has a deemphasis filter.

10. The device as claimed in one of patent claims 1 to 9, characterized in that the interference suppressor unit (2) carries out filtering in the time domain or frequency domain.

11. The device as claimed in one of patent claims 1 to 10, characterized in that the interference suppressor unit (2) has a comb filter with maximum attenuation values in each case in the vicinity of the fundamental frequency and the associated harmonics of the periodic interference signals.

12. The device as claimed in one of patent claims 1 to 11, characterized by an interference signal prefilter (C) for reducing the periodic interference signal in the useful signal (x(k)).

13. The device as claimed in one of patent claims 1 to 12, characterized in that the useful signal (x(k)) which has been subjected to interference is generated by an electret microphone.

14. The device as claimed in one of patent claims 1 to 13, characterized in that it is constructed in a wire-free telecommunications terminal.

15. The device as claimed in one of patent claims 1 to 13, characterized in that it is constructed in a hearing aid device.

16. The device as claimed in patent claim 15, characterized in that the hearing aid device constitutes a behind-the-ear device, an in-the-ear device, an in-the-canal device, a pocket device, a headset and/or an implant.

17. The device as claimed in one of patent claims 1 to 16, characterized in that the periodic interference signal constitutes a GSM signal and/or DECT signal and/or Bluetooth signal.

18. A method for suppressing periodic interference signals having the steps:

signal analysis is carried out in order to output a fault signal (d(k)) and associated coefficients (a_i) on the basis of a useful signal (x(k)) which has been subjected to interference;

interference suppression is carried out in order to generate a fault signal (d′

(k)) which has been subjected to interference suppression and has reduced periodic interference signals on the basis of the fault signal (d(k)); and

a signal synthesis is carried out in order to recover a useful signal (x*(k)) which has been subjected to interference suppression, on the basis of the fault signal (d′

(k)) which has been subjected to interference suppression, and the coefficients (a_i).

19. The method as claimed in patent claim 18, characterized in that, during the signal analysis, FIR filtering and/or IIR filtering are carried out in order to output a predictive fault signal (d(k)) and associated predictor coefficients (a_i) on the basis of a voice signal (x(k)), and the signal synthesis IIR filtering and/or IIR filtering are carried out in order to recover the useful signal (x*(k)) which has been subjected to interference suppression, on the basis of a predictive fault signal (d′

(k)) which has been subjected to interference suppression, and the predictor coefficients (a_i).

20. The method as claimed in patent claim 18 or 19, characterized in that a linear prediction is carried out during the signal analysis.

21. The method as claimed in patent claim 20, characterized in that the linear prediction constitutes a short-term prediction in a time domain of 20 to 400 milliseconds.

22. The method as claimed in one of patent claims 18 to 21, characterized in that, during the signal analysis, the coefficients (a_i) are determined by means of a Levinson-Durbin algorithm.

23. The method as claimed in one of patent claims 18 to 22, characterized by the step of carrying out high-pass filtering in order to filter the useful signal (x(k)) which has been subjected to interference, and in order to improve the coefficient calculation during the signal analysis.

24. The method as claimed in patent claim 23, characterized in that preemphasis filtering is carried out during the high-pass filtering.

25. The method as claimed in one of patent claims 23 or 24, characterized by the further step of low-pass filtering in order to filter the useful signal (x*(k)) which has been subjected to interference suppression, and in order to compensate the high-pass filtering.

26. The method as claimed in patent claim 25, characterized in that deemphasis filtering is carried out during the low-pass filtering.

27. The method as claimed in one of patent claims 18 to 26, characterized in that, during the interference suppression, filtering is carried out in the time domain or frequency domain.

28. The method as claimed in one of patent claims 18 to 27, characterized in that, during the interference suppression, comb filtering is carried out with maximum attenuation values in each case in the vicinity of the fundamental frequency and the associated harmonics of the periodic interference signals.

29. The method as claimed in one of patent claims 18 to 28, characterized by the additional step of carrying out interference signal prefiltering in order to reduce the periodic interference signal in the useful signal (x(k)).

30. The method as claimed in one of claims 18 to 29, characterized in that the useful signal (x(k)) which has been subjected to interference is generated by an electret microphone.

31. The method as claimed in one of patent claims 18 to 30, characterized in that it is carried out in a wire-free telecommunications terminal.

32. The method as claimed in one of patent claims 18 to 30, characterized in that it is carried out in a hearing aid device.

33. The method as claimed patent claim 32, characterized in that the periodic interference signal constitutes a GSM signal and/or DECT signal and/or Bluetooth signal.