SYSTEM AND METHOD FOR SPECTRAL GENERATION IN RADAR

• US 20040178951A1
• Filed: 03/07/2003
• Published: 09/16/2004
• Est. Priority Date: 03/13/2002
• Status: Active Grant

• Current Assignee: Raytheon Canada Limited (Rtx Corporation)
• Original Assignee: Raytheon Canada Limited (Rtx Corporation)
• Inventors: Dizaji, Reza, Ponsford, Tony
• Granted Patent: US 6,822,606 B2
• Time in Patent Office: Days
• US Class Current: 342/192
• CPC Class Codes: G01S 13/5244 Adaptive clutter cancellation specially adapted for airborne MTI, G01S13/5242
• CPC Class Codes: G01S 3/74 Multi-channel systems specially adapted for direction-finding, i.e. having a single antenna system capable of giving simultaneous indications of the directions of different signals systems in which the directions of different signals are determined sequentially and displayed simultaneously G01S3/04, G01S3/14
• CPC Class Codes: G01S 7/292 Extracting wanted echo-signals
• CPC Class Codes: G01S 7/2921 based on data belonging to one radar period
• CPC Class Codes: G01S 7/32 Shaping echo pulse signals; Deriving non-pulse signals from echo pulse signals

Abstract

This invention relates to a spectral generator and a spectral generation method for receiving pre-processed range-doppler-sensor data and generating at least one noise-reduced high-resolution spectrum therefrom. The spectral generator comprises a window generator that generates a window which defines a plurality of range-doppler cells. The spectral generator further comprises a covariance matrix calculator that is in communication with the window generator to receive the range-doppler-sensor data within the window and calculate a covariance matrix estimate for a range-doppler cell of interest in the window. The spectral generator also includes a spectral calculator that is in communication with the covariance matrix calculator to calculate a high-resolution spectral vector based on a location matrix and a noise subspace matrix estimate.

Claims

1. A spectral generator for radar comprising:

a) a window generator for generating a window, said window defining a plurality of range-doppler cells;

b) a covariance matrix calculator in communication with said window generator to for calculating a covariance matrix estimate for a range-doppler cell of interest in said window from pre-processed range-doppler-sensor data, wherein said covariance matrix estimate is obtained from covariance matrices calculated for at least a portion of said plurality of range-doppler cells within said window; and

,

c) a spectral calculator, in communication with said covariance matrix calculator, for calculating a high-resolution spectral vector based on a location matrix, and a noise subspace matrix estimate, said noise subspace matrix estimate being derived from said covariance matrix estimate.

2. The spectral generator of claim 1, wherein said covariance matrix calculator generates said covariance matrix estimate according to a weighted average of said covariance matrices, and said window generator defines weights used in said weighted average.

3. The spectral generator of claim 2, wherein said window generator varies the size of said window according to a radar mode of operation.

4. The spectral generator of claim 2, wherein said window generator varies the weights in said window according to a radar mode of operation.

5. The spectral generator of claim 2, wherein said window has at least two regions comprising a first region at an inner portion of said window and a second region surrounding said first region, wherein the weights in said second region are less than or equal to the weights in said first region.

6. The spectral generator of claim 2, wherein said window has three regions comprising a first region at an inner portion of said window, a second region surrounding said first region, and a third region surrounding said first region, wherein the weights in said second region are less than or equal to the weights in said first region and the weights in said third region are less than or equal to the weights in said second region.

7. The spectral generator of claim 1, wherein said spectral generator further comprises:

a) a subspace calculator in communication with said covariance matrix calculator for receiving said covariance matrix estimate and calculating said noise subspace matrix estimate based on at least a portion of the noise subspace of said covariance matrix estimate; and

,

b) a location matrix generator, in communication with said spectral calculator, for generating said location matrix to have at least one location vector defining an azimuth for which said high-resolution spectral vector is generated.

8. The spectral generator of claim 7, wherein said subspace calculator comprises:

a) an SVD module for receiving said covariance matrix estimate and performing a singular value decomposition thereon;

b) a dimension selector, in communication with said SVD module, for selecting a dimension b for said noise subspace; and

,

c) a noise subspace matrix calculator, in communication with said dimension selector and said SVD module, for calculating said noise subspace matrix estimate.

9. The spectral generator of claim 1, wherein said spectral calculator is adapted to calculate said high-resolution spectral vector according to

10. The spectral generator of claim 9, wherein a location vector A_θ

i in the location matrix A is given by A<sub>θ

i</sub>=[1e<sup>j2π

(d/λ

)sin(θ

i)</sup>e<sup>j4π

(d/λ

)sin(θ

i)</sup> . . . e<sup>j2π

(K−

G)d/λ

sin(θ

i)</sup>]^T where d is the spacing of a linear array that provides said range-doppler-sensor data, K is the number of sensors in the linear array, θ

_i is an azimuth angle and λ

is the wavelength of transmitted radar pulses.

11. The spectral generator of claim 9, wherein a location vector A_i(Φ

,θ

) in the location matrix A is given by

12. The spectral generator of claim 1, wherein said spectral generator further comprises a covariance matrix smoother, in communication with said covariance matrix calculator for smoothing said covariance matrix estimate by averaging two or more forward covariance matrices where each forward covariance matrix (C_Fi) is obtained from said covariance matrix estimate (C_est) according to C_Fi=C_est(i:

(K+1−

G+i−

1), i;

(K+1−

G+i−

1)).

13. The spectral generator of claim 1, wherein said spectral generator further comprises a covariance matrix smoother, in communication with said covariance matrix calculator for smoothing said covariance matrix estimate by averaging two or more backward covariance matrices where each backward covariance matrix (C_Bi) is obtained from said covariance matrix estimate (C_est) according to C_Bi=C*((K+1−

i);

(G+1−

i), (K+1−

i);

(G+1−

i)).

14. The spectral generator of claim 1, wherein said spectral generator further comprises a covariance matrix smoother, in communication with said covariance matrix calculator, for smoothing said covariance matrix estimate by averaging at least one forward covariance matrix and at least one backward covariance matrix where each forward covariance matrix (C_Fi) and each backward covariance matrix (C_Bi) is obtained from said covariance matrix estimate (C_est) according to C_Fi=C_est(i:

(K+1−

G+i−

1), i;

(K+1−

G+i−

1)) and C_Bi=C*((K+1−

i);

(G+1−

i), (K+1−

l);

(G+1−

i)).

15. The spectral generator of claim 1, wherein the spectral generator further comprises a noise suppression module, in communication with said covariance matrix calculator, for providing noise suppressed radar data instead of said pre-processed range-doppler-sensor data to said covariance matrix calculator, said noise suppression module estimating external interference in said pre-processed radar data and suppressing said external interference to produce said noise suppressed radar data.

16. The spectral generator of claim 15, wherein the noise suppression module comprises:

a) a first processing module for receiving said pre-processed radar data and producing matched radar data;

b) a second processing module for receiving said pre-processed radar data and producing mismatched radar data;

c) an adaptive beamformer in communication with said processing modules, for receiving a portion of matched and mis-matched radar data producing an estimate of external interference in said portion of matched radar data; and

,

d) a suppressor, in communication with said first processing module and said adaptive beamformer, for providing a portion of said noise suppressed radar data by suppressing said estimate of external interference from said portion of matched radar data.

17. A method of spectral generation for radar, said method comprising:

a) generating a window that defines a plurality of range-doppler cells;

b) calculating a covariance matrix estimate for a range-doppler cell of interest in said window from pre-processed range-doppler-sensor data, wherein said covariance matrix estimate is generated from covariance matrices calculated for at least a portion of said plurality of range-doppler cells within said window; and

,

c) calculating a high-resolution spectral vector based on a location matrix and a noise subspace matrix estimate, said noise subspace matrix estimate being derived from said covariance matrix estimate.

18. The spectral generation method of claim 17, wherein step (b) includes calculating said covariance matrix estimate based on a weighted average of said covariance matrices wherein said window defines weights used in said weighted average.

19. The spectral generation method of claim 18, wherein step (a) includes varying the size of said window according to a radar mode of operation.

20. The spectral generation method of claim 18, wherein step (a) includes varying the weights in said window according to a radar mode of operation.

21. The spectral generation method of claim 17, wherein step (c) includes:

d) calculating said noise subspace matrix estimate based on at least a portion of the noise subspace of said covariance matrix estimate; and

,

e) generating said location matrix having at least one location vector defining an azimuth for which said high-resolution spectral vector is generated.

22. The spectral generation method of claim 21, wherein step (d) includes:

i) performing a singular value decomposition of said covariance matrix estimate;

ii) selecting a dimension for said noise subspace; and

,

iii) calculating said noise subspace matrix estimate.

23. The spectral generation method of claim 17, wherein step (c) includes calculating said high-resolution spectral vector according to

24. The spectral generation method of claim 17, wherein said method further comprises smoothing said covariance matrix estimate by averaging two or more forward covariance matrices where each forward covariance matrix (C_Fi) is obtained from said covariance matrix estimate (C_est) according to C_Fi=C_est(i:

(K+1−

G+i−

1), i;

(K+1−

G+i−

1)).

25. The spectral generation method of claim 17, wherein said method further comprises smoothing said covariance matrix estimate by averaging two or more backward covariance matrices where each backward covariance matrix (C_Bi) is obtained from said covariance matrix estimate (C_est) according to C_Bi=C*((K+1−

i);

(G+1−

i), (K+1−

i);

(G+1−

i)).

26. The spectral generation method of claim 17, wherein said method further comprises smoothing said covariance matrix estimate by averaging at least one forward covariance matrix and at least one backward covariance matrix where each forward covariance matrix (C_Fi) and each backward covariance matrix (C_Bi) is obtained from said covariance matrix estimate (C_est) according to C_Fi=C_est(i:

(K+1−

G+i−

1), i;

(K+1−

G+i−

1)) and C_Bi=C*((K+1−

i);

(G+1−

i), (K+1−

i);

(G+1−

i)).

27. The spectral generation method of claim 17, wherein said method further comprises a step of suppressing noise to provide noise suppressed radar data instead of said pre-processed range-doppler-sensor data in step (b), said step of suppressing noise is effected by estimating external interference in said pre-processed radar data and suppressing said external interference to produce said noise suppressed radar data.

28. The spectral generation method of claim 27, wherein said step of suppressing noise comprises:

i) processing pre-processed radar data to produce matched radar data;

ii) processing pre-processed radar data to produce mis-matched radar data;

iii) selecting a portion of said matched and said mis-matched radar data and performing adaptive beamforming to produce an estimate of external interference in said portion of matched radar data; and

,

iv) producing a portion of said noise suppressed radar data by suppressing said estimate of external interference from said portion of matched radar data.

29. A noise suppression module for suppressing external interference in pre-processed radar data, said noise suppression module comprising:

a) a first processing module, in communication with said pre-processed radar data for producing matched radar data;

b) a second processing module, in communication with said pre-processed radar data for producing mis-matched radar data;

c) an adaptive beamformer in communication with said processing modules for receiving a portion of matched radar data and a portion of mis-matched radar data, and for producing an estimate of external interference in said portion of matched radar data; and

,

d) a suppressor, in communication with said first processing module and said adaptive beamformer, for providing a portion of noise suppressed radar data by suppressing said estimate of external interference from said portion of matched radar data.

30. The noise suppression module of claim 29, wherein the module further comprises a storage means, in communication with said suppressor for storing radar data.

31. The noise suppression module of claim 29, wherein said first processing module comprises:

a) a matched filter module, in communication with said pre-processed radar data for producing matched range-pulse-sensor data;

b) a first doppler processing module, in communication with said matched filter module for producing matched range-doppler-sensor data; and

,

c) a sensor selector, in communication with said first doppler processing module for selecting matched range-doppler data for a given sensor.

32. The noise suppression module of claim 29, wherein said second processing module comprises:

a) a mis-matched filter module, in communication with said pre-processed radar data for producing mis-matched range-pulse-sensor data;

b) a second doppler processing module, in communication with said mis-matched filter module for producing mis-matched range-doppler-sensor data; and

,

c) an auxiliary array selector in communication with said second doppler processing module for selecting at least a portion of said plurality of sensors to construct a virtual auxiliary sensor array and provide mis-matched range-doppler-sensor data for said virtual auxiliary sensor array.

33. The noise suppression module of claim 29, wherein said second processing module comprises:

a) a mis-matched filter module, in communication with said pre-processed radar data for producing mis-matched range-pulse-sensor data;

b) an auxiliary array selector, in communication with said mis-matched filter module for selecting at least a portion of said plurality of sensors to construct a virtual auxiliary sensor array and provide mis-matched range-pulse-sensor data for said virtual auxiliary sensor array; and

,

c) a second doppler processing module in communication with said auxiliary array selector for producing mis-matched range-doppler-sensor data for said virtual auxiliary sensor array.

34. The noise suppression module of claim 29, wherein said second processing module comprises:

a) an auxiliary array selector, in communication with pre-processed radar data for selecting at least a portion of said plurality of sensors to construct a virtual auxiliary sensor array and provide pre-processed radar data for said virtual auxiliary sensor array;

b) a mis-matched filter module, in communication with said auxiliary array selector for producing mis-matched range-pulse-sensor data for said auxiliary sensor array; and

,

c) a second doppler processing module, in communication with said mis-matched filter module for producing mis-matched range-doppler-sensor data for said auxiliary sensor array.

35. The noise suppression module of claim 29, wherein said adaptive beamformer comprises:

a) an autocorrelation matrix calculator, in communication with said second processing module for calculating the autocorrelation matrix of said portion of mis-matched radar data;

b) a matrix inverter, in communication with said autocorrelation matrix calculator for providing an inverse autocorrelation matrix;

c) a cross-correlator, in communication with said processing modules for providing the cross-correlation of said portion of mis-matched radar data and said portion of matched radar data;

d) a weight calculator, in communication with said matrix inverter and said cross-correlator for calculating an adaptive weight vector based on said inverse autocorrelation matrix and said cross-correlation; and

,

e) an auxiliary beam generator, in communication with said second processing module and said weight calculator for providing said estimate of external interference based on said adaptive weight vector and said portion of mis-matched radar data.

36. The noise suppression module of claim 29, wherein said suppressor comprises:

a) a subtractor, in communication with said first processing module and said adaptive beamformer for subtracting said estimate of external interference from said portion of matched radar data to provide a subtraction signal; and

,

b) a minimizer, in communication with said subtractor and said first processing module for providing a portion of said noise suppressed radar based on said subtraction signal and said portion of matched radar data.

37. The noise suppression module of claim 29, wherein said noise suppression module further comprises an ordered-statistics filter module, in communication with said first processing module for producing ordered-statistics filtered matched radar data.

38. The noise suppression module of claim 37, wherein said ordered-statistics filter module performs median filtering.

39. The noise suppression module of claim 37, wherein said adaptive beamformer comprises:

a) an autocorrelation matrix calculator, in communication with said second processing module for calculating the autocorrelation matrix of said portion of mis-matched radar data;

b) a matrix inverter, in communication with said autocorrelation matrix calculator for providing an inverse autocorrelation matrix;

c) a cross-correlator, in communication with said ordered-statistics filter module and said second processing module for providing the cross-correlation of said portion of mis-matched radar data and a portion of ordered-statistics filtered matched radar data;

d) a weight calculator, in communication with said matrix inverter and said cross-correlator for calculating an adaptive weight vector based on said inverse autocorrelation matrix and said cross-correlation; and

,

e) an auxiliary beam generator, in communication with said second processing module and said weight calculator for producing said estimate of external interference based on said adaptive weight vector and said portion of mis-matched radar data.

40. A method of suppressing external interference in pre-processed radar data, said method comprising:

a) processing said pre-processed radar data to produce matched radar data;

b) processing said pre-processed radar data to produce mis-matched radar data;

c) selecting a portion of said matched radar data and a portion of said mis-matched radar data and performing adaptive beamforming to produce an estimate of the external interference in said portion of matched radar data; and

,

d) producing a portion of noise suppressed radar data by suppressing said estimate of external interference from said portion of matched radar data.

41. The method of claim 40, wherein said method further comprises:

e) storing said portion of noise suppressed radar data.

42. The method of claim 40, wherein step (a) comprises:

i) match filtering said radar data to produce matched range-pulse-sensor data;

ii) doppler processing said matched range-pulse-sensor data to produce matched range-doppler-sensor data; and

,

iii) selecting a portion of matched range-doppler data for a given sensor.

43. The method of claim 40, wherein step (b) comprises:

i) mis-match filtering said pre-processed radar data to produce mis-matched range-pulse-sensor data;

ii) doppler processing said mis-matched range-pulse sensor data to produce mis-matched range-doppler-sensor data; and

,

iii) constructing a virtual auxiliary sensor array from at least a portion of said plurality of sensors to provide mis-matched range-doppler-sensor data for said virtual auxiliary sensor array.

44. The method of claim 40, wherein step (b) comprises:

i) mis-match filtering said pre-processed radar data to produce mis-matched range-pulse-sensor data;

ii) constructing a virtual auxiliary sensor array from at least a portion of said plurality of sensors to provide mis-matched range-pulse-sensor data for said virtual auxiliary sensor array; and

,

iii) doppler processing said mis-matched range-pulse sensor data from said virtual auxiliary sensor array to produce mis-matched range-doppler-sensor data for said virtual auxiliary sensor array.

45. The method of claim 40, wherein step (b) comprises:

i) constructing a virtual auxiliary array from at least a portion of said plurality of sensors to provide pre-processed radar data for said virtual auxiliary sensor array;

ii) mis-match filtering said pre-processed radar data from said virtual auxiliary sensor array to produce mis-matched range-pulse-sensor data for said auxiliary sensor array; and

,

iii) doppler processing said mis-matched range-pulse-sensor data to produce mis-matched range-doppler-sensor data for said virtual auxiliary sensor array.

46. The method of claim 40, wherein step (c) comprises:

i) autocorrelating said portion of mis-matched radar data to produce an autocorrelation matrix;

ii) inverting said autocorrelation matrix to produce an inverse autocorrelation matrix;

iii) cross-correlating said portion of mis-matched radar data and said portion of matched radar data to produce a cross-correlation vector;

iv) calculating an adaptive weight vector; and

,

v) generating an auxiliary beam to produce said estimate of external interference.

47. The method of claim 40, wherein step (d) comprises:

i) subtracting said estimate of external interference from said portion of matched radar data to produce a subtraction signal; and

,

ii) producing said portion of noise suppressed matched radar data based on said subtraction signal and said portion of matched radar data.

48. The method of claim 40, wherein said method further comprises performing ordered-statistics filtering on said matched radar data to produce ordered-statistics filtered matched radar data.

49. The method of claim 48, wherein said ordered-statistics filtering comprises median filtering.

50. The method of claim 48, wherein step (c) comprises:

i) autocorrelating said portion of mis-matched radar data to produce an autocorrelation matrix;

ii) inverting said autocorrelation matrix to produce an inverse autocorrelation matrix;

iii) cross-correlating said portion of mis-matched radar data and a portion of ordered-statistics filtered matched radar data to produce a cross-correlation vector;

iv) calculating an adaptive weight vector; and

,

v) generating an auxiliary beam to produce said external interference estimate.