Method and circuit arrangement for processing a signal containing interference

• US 6,754,293 B1
• Filed: 05/30/1996
• Issued: 06/22/2004
• Est. Priority Date: 06/01/1995
• Status: Expired due to Fees

• Current Assignee: Nokia Mobile Phones UK Limited (Nokia Corporation)
• Original Assignee: Nokia Mobile Phones UK Limited (Nokia Corporation)
• Inventors: Henriksson, Jukka A., Raivio, Kimmo J.
• Primary Examiner(s): TSE, YOUNG TOI
• Application Number:
  US08/655,633
• Time in Patent Office: 2,945 Days
• Field of Search: 375/232, 375/233, 375/340, 375/341, 375/346, 375/348, 375/350, 708/319, 708/322, 708/323
• US Class Current: 375/346
• CPC Class Codes: H03H 21/0012 Digital adaptive filters

Abstract

The invention relates to a method and circuit arrangement for compensating for an interference component included in a signal in a digital transmission system such as a mobile communication system, radio system or television system. In the method according to the invention a decision error representing an interfering signal is classified using an adaptive method, and a table is drawn up of interference estimates corresponding to classified decision errors. So it is obtained on the basis of previous decision errors an interference estimate corresponding to the next signal sample and that estimate is subtracted from the signal sample before the current decision. The adaptive method is preferably a neural method such as a self-organizing map. With the method according to the invention it is possible to compensate for an interference component in a signal regardless of the interference source assuming that the interfering signal has internal correlation. If required, it is also possible to include in the method according to the invention equalization for delayed payload signal components.

Claims

1. A method for compensating for an interfering signal included in an input signal (r_k+1) carrying a digital payload signal to obtain an interference-compensated signal (z_k+1), comprising the steps of:

producing an interference estimate (ê

_k+1) for an interfering signal sample value;

predicting from said interference estimate (ê

_k+1) a next sample value for the interfering signal;

subtracting said interference estimate (ê

_k+1) from a signal sample (y_k+1) of an uncompensated signal derived from said input signal (r_k+1);

making a decision (s_k+1) on a sample value of the interference-compensated signal (z_k+1);

wherein said producing step includes a step of adaptively forming said interference estimate (ê

_k+1) as a function of at least one previous decision value (s_k) of said decision and at least one previous signal sample (y_k) of said uncompensated signal.

2. The method of claim 1, wherein said adaptive forming step is accomplished by a neural network.

3. The method of claim 1, wherein a previous sample (ε

_k) of said interfering signal is measured by calculating a difference of said previous decision and the corresponding uncompensated signal sample (y_k).

4. The method of claim 1, further comprising the steps of equalizing and delaying said input signal (r_k+1) before the subtraction of the interference estimate (ê

_k+1) from the uncompensated signal (y_k+1) to accomplish said interference compensation.

5. The method of claim 1, wherein, in said step of producing said interference estimate (ê

_k+1), said producing employs the output value (s_k+1) corresponding to at least one intermediate decision (s_k2) of said decision making step.

6. The method of claim 1, further comprising a step of transmitting the interference-compensated signal (z_k+1) in a transmission system conveying the interference-compensated signal as a digitally modulated signal suitable for use in a mobile communication system, radio system or television system.

7. A method for compensating for an interfering signal included in an input signal (r_k+1) carrying a digital payload signal to obtain an interference-compensated signal (z_k+1), comprising the steps of:

producing an interference estimate (ê

_k+1) for an interfering signal sample value;

predicting from said interference estimate (ê

_k+1) a next sample value for the interfering signal;

subtracting paid interference estimate (ê

_k+1) from a signal sample (y_k+1) of an uncompensated signal derived from said input signal (r_k+1);

making a decision (s_k+1) on a sample value of the interference-compensated signal (z_k+1);

wherein said producing step includes a step of adaptively forming said interference estimate (ê

_k+1) as a function of at least one previous decision value (s_k) of said decision and at least one previous signal sample (y_k) of said uncompensated signal; and

interference samples (ε

_k, ε

_k+1, . . . ε

<sub>k−

(N−

1)</sub>) of said interfering signal are matched to a self-organizing map to convert, or classify, the interference samples into map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>).

8. The method of claim 7, wherein the interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>) are converted into the map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>) in the following stages;

one of the interference samples (ε

_k) is fed into the self-organizing map,

the interference sample is matched to one of the map points (m_k) having the shortest Euclidian distance to the interference sample,

the map point (m_k) is fed into a first shift register to produce a stored sequence of map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>), and

updating the map.

9. The method of claim 7, wherein at least part of the combinations (M_(N−

1)j, . . . , M_Ij, M_oj) are map points of the self organizing map and are logically associated with corresponding estimates (ê

_j) stored in an estimate register.

10. The method of claim 9, wherein the estimates (ê

_j) in the register correspond with the map points (m_k, m_k+1, . . . , m<sub>k−

(N−

1)</sub>) of the map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>) obtained from the interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>) is used as a next interference estimate (ê

_k).

11. The method of claim 10, further comprising a step of updating the value of the estimate (ê

_j+1) used as the interference estimate before selecting the next interference estimate.

12. The method of claim 11, wherein the updating of the estimate value is done by using substantially a gradient method.

13. The method of claim 12, wherein the estimate register provides a self-organizing map (SOM), and the estimate values of the neighborhood of the selected estimate (ê

_j+1) are updated using substantially a SOM algorithm.

14. The method of claim 12, wherein the updating of the self-organizing map (SOM) is accomplished by using substantially the SOM algorithm.

15. The method of claim 7, wherein conversion of the interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>) into the map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>) is done in the following stages;

N last interference samples of a set of the interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>) are stored in a second shift register,

an interference sample vector for the set of the interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>) is formed of the interference samples stored in the second shift register, and the vector is fed into the self organizing map, the self organizing map being N-dimensional,

the interference sample vector fed into the map is matched to a map point having the shortest Euclidian distance to the interference sample vector in order to convert the interference sample vector into a classification result vector of the map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>), and

updating the map.

16. The method of claim 15, wherein at least part of the points of the self-organizing map are logically associated with corresponding estimates (ê

_j) in an estimate register.

17. The method of claim 16, wherein a further estimate (ê

_j+1) in the estimate register corresponding to the map point of the classification result vector (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>) obtained from the last interference samples of the set of the interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>) is used as a next interference estimate (ê

_k+1).

18. A circuit arrangement for compensating for an interfering signal included in an input signal (r_k) carrying a digital payload signal, including:

decision means, and means for producing an interference estimate (ê

_k+1);

means for subtracting said interference estimate from a signal sample (y_k+1) of an uncompensated signal derived from said input signal;

a decision circuit for making a decision (s_k+1) on an interference-compensated signal (z_k+1);

wherein said means for producing the interference estimate (ê

_k+1) comprises means for adaptively determining the interference estimate (ê

_k+1) on the basis of at least one previous uncompensated signal sample (y_k) and at least one previous decision (s_k) of said decision means; and

said adaptive determining means comprises means for subtracting said interference estimate (ê

_k+1) from said previous uncompensated signal sample (y_k) to provide an input signal to said decision means.

19. The circuit arrangement of claim 18, further comprising;

means for measuring an interference sample (ε

_k),

means for adaptively classifying the interference sample (ε

_k), and converting the interference sample (ε

_k) into a map point (m_k), and

means for determining the next one of an interference estimate (ê

_k+1) as a function of the map point (m_k).

20. The circuit arrangement of claim 18, further comprising a FIR filter for equalizing delayed payload signal components before compensation of the interference signal.

21. The circuit arrangement of claim 18, further comprising a decision feedback equalizer for equalizing delayed payload signal components, said decision feedback equalizer comprising:

means for forming the difference of a decision circuit input and output for a decision error,

a FIR filter for forming a weighted sum of previous decisions constituting an equalization value on the basis of the decision error and a decision value, and

means for subtracting the decision value from the input signal before compensation of the interfering signal and a making of the decision.

22. The circuit arrangement of claim 18, further comprising means for forming at least one intermediate decision (s_k2) and means for forming the interference estimate (ê

_k+1)as a function of said at least one intermediate decision (s_k2).

23. The circuit arrangement of claim 18, wherein said circuit arrangement is part of a transmission system conveying a digitally modulated signal;

suitable for use as a mobile communication system, radio system, or television system.

24. A circuit arrangement for compensating for an interfering signal included in an input signal (r_k) carrying a digital payload signal, including:

decision means, and means for producing an interference estimate (ê

_k+1);

means for subtracting said interference estimate from a signal sample (y_k+1) of an uncompensated signal derived from said input signal;

a decision circuit for making a decision (s_k+1) on an interference-compensated signal (z_k+1);

means for measuring an interference sample (ε

_k);

means for adaptively classifying the interference sample (ε

_k), and converting the interference sample (ε

_k) into a map point (m_k); and

means for determining the next one of an interference estimate (ê

_k+1) as a function of the ma pint (m_k);

wherein said means for producing the interference estimate (ê

_k+1) comprises means for adaptively determining the interference estimate (ê

_k+1) on the basis of at least one previous uncompensated signal sample (y_k) and at least one previous decision (s_k); and

wherein said means for adaptively classifying the interference sample (ε

_k) include a self organizing map.

25. The circuit arrangement of claim 24 further comprising a shift register for storing the last interference samples of a set of interference samples (ε

_k, ε

<sub>k−

1</sub>, . . . ε

<sub>k−

(N−

1)</sub>), and wherein the self-organizing map includes dimensions for two or more interference samples stored in the second shift register.

26. A circuit arrangement for compensating for an interfering signal included in an input signal (r_k) carrying a digital payload signal, including:

decision means, and means for producing an interference estimate (ê

_k+1);

means for subtracting said interference estimate from a signal sample (y_k+1) of an uncompensated signal derived from said input signal;

a decision circuit for making a decision (s_k+1) on an interference-compensated signal (z_k+1);

means for measuring an interference sample (ε

_k);

means for, adaptively classifying the interference sample (ε

_k), and converting the interference sample (ε

_k) into a map point (m_k); and

means for, determining the next one of an interference estimate (ê

_k+1) as a function of the map point (m_k);

wherein said means for producing the interference estimate ê

_k+1) comprises means for adaptively determining the interference estimate ê

_k+1) on the basis of at least one previous uncompensated signal sample (y_k) and at least one previous decision (s_k); and

a shift register for storing a succession of map points (m_k, m<sub>k−

1</sub>, . . . , m<sub>k−

(N−

1)</sub>) for determining the interference estimate (ε

_k+1).

27. A circuit arrangement for compensating for an interfering signal included in an input signal (r_k) carrying a digital payload signal, including:

decision means, and means for producing an interference estimate (ê

_k+1);

means for subtracting said interference estimate from a signal sample (y_k+1) of an uncompensated signal derived from said input signal;

a decision circuit for making a decision (s_k+1) on an interference-compensated signal (z_k+1);

means for measuring an interference sample (ε

_k);

means for adaptively classifying the interference sample (ε

_k), and converting the interference sample (ε

_k) into a map point (m_k); and

means for determining the next one of an interference estimate (ê

_k+1) as a function of the map point (m_k);

wherein said means for producing the interference estimate (ê

_k+1) comprises means for adaptively determining the interference estimate (ê

_k+1) on the basis of at least one uncompensated signal sample (y_k) and at least one previous decision (s_k);

an estimate register with estimate values (ê

_j) therein, and

means for selecting the estimate values (ê

_j) from the estimate register on the basis of one or more previous map points (m_k) for the interference estimate (ê

_k+1).

28. The circuit arrangement of claim 27, further comprising means for updating the estimate values (ê

_j) stored in the estimate register.

29. A method for compensating for an interfering signal included in an input signal (r_k+1) carrying a digital payload signal to obtain an interference-compensated signal (z_k+1), comprising the steps of:

producing an interference estimate (ê

_k+1) for an interfering signal sample value;

predicting from said interference estimate (ê

_k+1) a next sample value for the interfering signal;

subtracting said interference estimate (ê

_k+1) from a signal sample (y_k+1) of an uncompensated signal derived from said input signal (r_k+1);

making a decision (s_k+1) on a sample value of the interference-compensated signal (z_k+1);

wherein said producing step includes a step of adaptively forming said interference estimate (ê

_k+1) as a function of at least one previous decision value (s_k) of said decision and at least one previous signal sample (y_k) of said uncompensated signal; and

storing points of a map based on decisions of said decision-making step, and updating a set of said points as a self-organizing map for producing values of the interference estimate (ê

_k+1).

30. A circuit arrangement for compensating for an interfering signal included in an input signal (r_k) carrying a digital payload signal, including:

decision means, and means for producing an interference estimate (ê

_k+1);

means for subtracting said interference estimate from a signal sample (y_k+1) of an uncompensated signal derived from said input signal;

a decision circuit for making a decision (s_k+1) on an interference-compensated signal (z_k+1);

wherein said means for producing the interference estimate (ê

_k+1) comprises means for adaptively determining the interference estimate (ê

_k+1) on the basis of at least one uncompensated signal sample (y_k) and at least one previous decision (s_k) of said decision means; and

said adaptive determining means comprises means for subtracting said interference estimate (ê

_k+1) from said uncompensated signal sample (y_k) to provide an input signal to said decision means, and map means operative on output signals of said decision means as a self-organizing map for producing values of the interference estimate (ê

_k+1).