Compensating for distortion in a communication channel

• US 4,922,506 A
• Filed: 01/11/1988
• Issued: 05/01/1990
• Est. Priority Date: 01/11/1988
• Status: Expired due to Term

• Current Assignee: Intersil Americas LLC (Renesas Electronics Corporation)
• Original Assignee: Sicom Corporation
• Inventors: McCallister, Ronald D., Shearer, Daniel D. III
• Primary Examiner(s): CANGIALOSI, SALVATORE A
• Application Number:
  US07/142,707
• Time in Patent Office: 841 Days
• Field of Search: 375/1, 380/34, 324/57, 324/58.5 A, 364/553, 364/554, 364/802, 364/485, 364/724.02, 364/724.01
• US Class Current: 375/142
• CPC Class Codes: H04B 1/707 using direct sequence modulation
• CPC Class Codes: H04L 27/2278 using correlation techniques, e.g. for spread spectrum signals

Abstract

This disclosure relates to a technique for estimating the characteristics of a propagation channel by processing a cyclostationary signal which has passed through it. The technique extends to the use of this estimate to reduce the impact of the channel distortion on the recovered signal. Unique features of this technique include its ability to function even when strong interference is present with the cyclostationary signal of interest.

Claims

1. In the method of estimating transfer characteristics of a propagation medium from a cyclostationary signal received therefrom comprising the steps of:

generating a plurality of estimator signals only from the received cyclostationary signal, each of which is representative of a spectrally discrete portion of a correlation of said received signal; and

using said plurality of estimator signals to provide an estimate of the transfer characteristics of the medium.

2. In the method of claim 1 wherein the step of generating said estimator signals includes the step of generating a correlation signal proportional to the correlation of said received signal with a known correlation delay.

3. In the method of claim 1 wherein the step of generating said estimator signals includes the step of filtering out noise that is spectrally diffuse.

4. In the method of claim 2 wherein the step of generating said correlation signal includes the step of delaying said received signal by said correlation delay to provide a delayed received signal, said correlation signal being proportional to the product of the amplitude of said delayed received signal and the amplitude of said received signal.

5. In the method of claim 2 wherein the step of generating said estimator signals includes the step of down converting to DC a spectrally discrete component of said correlation signal that is in phase with a harmonic of the chip rate of said received signal.

6. In the method of claim 2 wherein the step of generating said estimator signals includes the step of demodulating a spectrally discrete component of said correlation signal that is in quadrature with a harmonic of the received signal.

7. The method of claim 1 wherein said estimate is provided by the following equation:

##EQU14## Where;

C.sub.(1)n (τ

_k) is the discrete portion of the complex correlation of said received signal, with a correlation delay

T is the clock period of the received signal;

n is an index of harmonics of the clock rate;

w is radian frequency;

w_n is the radian frequency of a harmonic of the clock rate;

τ

s is a timing synchronization error;

P(w) is the Fourier transform of the baseband signal;

P*(w-w_n) is the complex conjugate of P(w-w_m);

H_m (w) is the transfer function of the model; and

H_m *(w-w_n) is the complex conjugate of H_m (w-w_n).

8. The method of claim 1 wherein said estimate is given as H_m (w) which is defined as follows:

##EQU15## Where;

H_m(w) is the m^th iterative estimate;

ε

m is the m^th iterative value of the i^th numerator coefficient;

λ

im is the m^th iterative value of the i^th denominator coefficient; and

N is the number of coefficients ε

i or λ

i utilized in forming said estimate.

9. In the method of claim 1 wherein the received signal is multiplied by an estimated radio frequency carrier signal including the steps of:

generating an in-phase received signal proportional to the product of the amplitude of an in-phase component of an estimated carrier signal and the amplitude of said received signal; and

generating a quadrature received signal proportional to the product of the amplitude of a quadrature component of said estimated carrier signal and the amplitude of said received signal.

10. In the method of claim 9 wherein the step of generating said estimator signals includes the step of generating a signal proportional to an in-phase part of the complex correlation of said received signal with a known correlation delay.

11. In the method of claim 10 wherein the step of generating said in-phase complex correlation signal includes the step of delaying said in-phase received signal by said correlation delay to provide a delayed in-phase received signal, said in-phase complex correlation signal being proportional to the product of the amplitude of said delayed in-phase received signal and the amplitude of said in-phase received signal.

12. In the method of claim 9 wherein the step of generating said estimator signals includes the step of generating a signal proportional to a quadrature part of the complex correlation of said received signal with a known correlation delay.

13. In the method of claim 12 wherein the step of generating said quadrature complex correlation signal includes the step of delaying said in-phase received signal by said correlation delay, to provide a delayed in-phase received signal, said quadrature correlation signal being proportional to the product of the amplitude of said delayed in-phase received signal and the amplitude of said quadrature received signal.

14. In the method of claim 9 wherein the carrier frequency of said received signal and the frequency of said estimated carrier signal are of the same frequency.

15. Apparatus for providing an estimate of transfer characteristics of a propagation medium from a cyclostationary signal received therefrom comprising, in combination:

means for generating a plurality of estimator signals only from the received cyclostationary signal, each of which is representative of a spectrally discrete portion of a correlation of said received signal; and

means for using said plurality of estimator signals to provide an estimate of the transfer characteristics of the medium.

16. The apparatus of claim 15 wherein said generating means includes means for generating a correlation signal proportional to the correlation of said received signal with a known correlation delay.

17. The apparatus of claim 15 wherein said generating means includes means for filtering out noise that is spectrally diffuse.

18. The apparatus of claim 16 wherein said means for generating said correlation signal comprises:

a time delay network having an input where said received signal is applied, said network providing an output substantially the same as an input signal applied thereto, but delayed by a known time; and

a multiplier unit having one of two inputs for said received signal, the output of said network being connected to the other input, said correlation signal being provided at the output of said multiplier with a correlation delay substantially equal to said known time.

19. The apparatus of claim 16 wherein said generating means includes means for down-converting to DC a spectrally discrete component of said correlation signal that is in-phase with a harmonic of a clock rate of said received signal.

20. The apparatus of claim 16 wherein said generating means includes means for down converting to DC a spectrally discrete component of said correlation signal that is in quadrature with a harmonic of a clock rate of said received signal.

21. The apparatus of claim 15 wherein said estimate is provided by the following equation:

##EQU16## Where;

C.sub.(1)n (τ

_k) is the discrete portion of the complex correlation of said received signal, with a correlation delay τ

k

T is the clock period;

n is an index of harmonics of the clock rate;

w is radian frequency;

τ

_s is a timing synchronization error;

P(w) is the Fourier transform of the baseband signal;

P*(w-w_n) is the complex conjugate of P(w-w_n);

H_m *(w-w_n) is the complex conjugate of Hm(w-w_n);

H_m (w) is the estimate of the transfer function of the medium.

22. The apparatus of claim 15 wherein said estimate is given as H_m (w) which is defined as follows:

##EQU17## Where;

H_m (w) is the m^th iterative estimate;

ε

m is the m^th iterative value of the i^th numerator coefficient;

λ

im is the m^th iterative value of the i^th denominator coefficient; and

N is the number of coefficients ε

i or λ

i utilized in forming said estimate.

23. The apparatus of claim 15 wherein the received signal is multiplied by an estimated radio frequency carrier signal;

means for generating an in-phase received signal proportional to the product of the amplitude of an in-phase component of an estimated carrier signal and the amplitude of a signal received from said medium; and

means for generating a quadrature received signal proportional to the product of the amplitude of a quadrature component of said estimated carrier signal and the amplitude of said received signal.

24. The apparatus of claim 23 wherein said generating means includes means for generating a signal proportional to an in-phase part of the complex correlation of said received signal with a known time delay.

25. The apparatus of claim 24 wherein said generating means comprises:

a time delay network having an input where said in-phase received signal is applied, said network providing an output substantially the same as an input signal applied thereto, but delayed by a known time; and

a multiplier unit having one of two inputs where said in-phase received signal is applied, the output of said delay network being connected to the other input of said multiplier, the output of said multiplier being an in-phase portion of the complex correlation of said received signal with a correlation delay of said known time.

26. The apparatus of claim 25 wherein said generating means comprises:

a multiplier unit having one of two inputs where said quadrature received signal is applied, the output of said delay network being connected to the other input of said multiplier, the output of said multiplier being a quadrature portion of the complex correlation of said received signal with a correlation delay of said known time.

27. The apparatus of claim 26 wherein said generating means comprises:

means for down converting to DC a part of said in-phase portion of said complex correlation signal that is in-phase with a harmonic of the clock rate of said received signal;

means for down converting to DC a part of said in-phase portion of said complex correlation signal that is in quadrature with said harmonic of the clock rate of said received signal;

means for down converting to DC a part of said quadrature portion of said complex correlation signal that is in-phase with said harmonic of the clock rate of said received signal; and

means for down converting to DC a part of said quadrature portion of said complex correlation signal that is in quadrature with said harmonic of the clock rate of said received signal.

28. The apparatus of claim 27 wherein at least one of said down converting means comprises a multiplying unit.

29. In the method of recovering a baseband signal from a spread spectrum signal received from a propagation medium, said spread spectrum signal being generated by modulating the baseband signal with a reference spreading signal, the steps of:

generating a plurality of estimator signals, each of which is representative of a spectrally discrete portion of a correlation of said received signal;

providing a representation of an assumed set of values of coefficients of an equation of the transfer function of a model that is an estimate of the transfer characteristics of the medium;

generating a plurality of model signals, respectively corresponding to said estimator signals, in accordance with a relationship between a discrete portion of the correlation of the output of said model and the transfer function of said model;

iteratively adjusting said values to reduce differences between said estimator signals and said respectively corresponding model signals; and

providing a local spreading signal synchronized to said reference spreading signal, that is modified in accordance with the estimated transfer function.

30. In the method of claim 29 wherein the step of generating said estimator signals includes the step of generating a correlation signal proportional to the correlation of said received signal with a known correlation delay.

31. In the method of claim 29 wherein the step of generating said estimator signals includes the step of filtering out noise that is spectrally diffuse.

32. In the method of claim 30 wherein the step of generating said correlation signal includes the step of delaying said received signal by said correlation delay to provide a delayed received signal, said correlation signal being proportional to the product of the amplitude of said delayed received signal and the amplitude of said received signal.

33. In the method of claim 30 wherein the step of generating said estimator signals includes the step of down converting to DC a spectrally discrete component of said correlation signal that is in phase with a harmonic of the chip rate of said spreading signal.

34. In the method of claim 30 wherein the step of generating said estimator signals includes the step of demodulating a spectrally discrete component of said correlation signal that is in quadrature with a harmonic of the spreading signal.

35. The method of claim 29 wherein said relationship is given as:

##EQU18## Where;

C.sub.(1)n (τ

_k) is the discrete portion of the complex correlation of the model, with a correlation delay, τ

_k ;

T is the chip period of said received signal;

m is an index of harmonics of the chip rate;

w is radian frequency;

w_n is the radian frequency of a harmonic of the chip rate;

τ

_s is a timing synchronization error;

P(w) is the Fourier transform of the baseband signal;

P*(w-w_n) is the complex conjugate of P(w-w_n);

H_m (w) is the estimate of the transfer function of the medium; and

H_m *(w-w_n) is the complex conjugate of H_m (w-w_n).

36. The method of claim 29 wherein said transfer function of said model is given as:

##EQU19## Where;

H_m(w) is the m^th iterative estimate

ε

im is the m^th iterative value of the i^th numerator coefficient;

λ

im is the m^th iterative value of the i^th denominator coefficient; and

N is the number of coefficients ε

i or λ

i utilized in forming said estimate.

37. In the method of claim 29 wherein the received signal is multiplied by an estimated radio frequency carrier signal, including the steps of:

generating an in-phase spread spectrum signal proportional to the product of the amplitude of an in-phase component of an estimated carrier signal and the amplitude of a signal received from said medium; and

generating a quadrature spread spectrum signal proportional to the product of the amplitude of a quadrature component of said estimated carrier signal and the amplitude of said received signal.

38. In the method of claim 37 wherein the step of generating said estimator signals includes the step of generating a signal proportional to an in-phase part of the complex correlation of said received signal with a known correlation delay.

39. In the method of claim 38 wherein the step of generating said in-phase complex correlation signal includes the step of delaying said in-phase spread spectrum signal by said correlation delay to provide a delayed in-phase spread spectrum signal, said in-phase complex correlation signal being proportional to the product of the amplitude of said delayed in-phase spread spectrum signal and the amplitude of said in-phase spread spectrum signal.

40. In the method of claim 37 wherein the step of generating said estimator signals includes the step of generating a signal proportional to a quadrature part of the complex correlation of said received signal with a known correlation delay.

41. In the method of claim 40 wherein the step of generating said quadrature complex correlation signal includes the step of delaying said in-phase spread spectrum signal by said correlation delay to provide a delayed in-phase spread spectrum signal, said quadrature correlation signal being proportional to the product of the amplitude of said delayed in-phase spread spectrum signal and the amplitude of said quadrature spread spectrum signal.

42. In the method of claim 37 wherein the carrier frequency of said received signal and the frequency of said estimated carrier signal are of the same frequency.

43. A matched filter for received spread signals that are demodulated by sequences provided by first and second spread sequence generators, respectively, to provide first and second spread baseband signals, said sequences being of the same chip rate but differing from each other, comprising:

generating means for providing an in-phase RF carrier signal and a quadrature RF carrier signal that are phase shifted 90°

from each other;

means for downconverting said received spread signals with said in-phase and quadrature RF signals, respectively, to provide a pair of first and second baseband spread spectrum signals, said downconverting means being adapted for coupling to a medium through which said received spread spectrum signals are propagated;

estimating means coupled to said medium for generating a representation of the coefficients of an equation of a model of the transfer characteristics of said medium;

simulation means for providing first and second spreading signals that are modified by said model; and

recovery means for providing first and second recovered baseband signals in response to said spreading signals and said first and second baseband spread spectrum signals.

44. The matched filter of claim 43 wherein said generating means comprises:

an RF carrier generator synchronized to said received signal; and

a hybrid that has its input connected to the output of said carrier generator, and outputs that provide an inphase synchronized carrier signal and a quadrature synchronized carrier signal.

45. The matched filter of claim 43 wherein the signal coupled to said medium for propagation therethrough is in accordance with a relationship which is given as:

space="preserve" listing-type="equation">M.sub.I (t)=d.sub.I (t) S .sub.I (t ) cos (w.sub.c t)+d.sub.Q (t)S.sub.Q (t) sin (w.sub.c t)

Where;

M_I (t) represents the amplitude of said signal coupled to said medium;

d_I (t) represents the amplitude of the first baseband signal;

S_I (t) represents the amplitude of the output of the first spread sequence generator;

W_c represents the carrier frequency;

d_Q (t) represents the amplitude of the second baseband signal; and

S_Q (t) represents the amplitude of the output of the second spread sequence generator.

46. The matched filter of claim 43 wherein said received signal is in accordance with a relationship which is given as:

space="preserve" listing-type="equation">Mo(t)=d.sub.I (t)S.sub.I1 (t) cos (W.sub.c t)+d.sub.Q (t)S.sub.Q1 (t) sin (W.sub.c t)+d.sub.I (t)S.sub.I2 (t ) sin (W.sub.c t)-d.sub.Q (t) S.sub.Q2 (t) cos (W.sub.c t)

Where;

Mo(t) represents said received signal;

the carat ( ) notation is indicative of distortion that may be caused by said member;

S_I1 (t) is the amplitude of a component of said received signal provided in reference to said first spread baseband signal;

d_I (t) is the amplitude of said first baseband signal;

S_Q2 (t) is the amplitude of a component of said received signal provided in response to said second spread baseband signal; and

d_Q (t) is the amplitude of said second baseband signal.

47. The matched filter of claim 43 wherein said estimating means includes:

means for generating an estimated carrier signal that has a frequency that is an estimate of said carrier frequency;

means for down shifting said received signal with said estimated carrier signal, thereby providing a frequency shifted spread spectrum signal comprising a frequency shifted in-phase spread spectrum signal and a frequency shifted quadrature spread spectrum signal; and

means for approximating the transfer function of said medium.

48. The matched filter of claim 47 wherein said means for down shifting said received signal comprises:

a hybrid with the input thereof connected to said estimated carrier generating means; and

a multiplier unit having a pair of inputs coupled to the output of said medium and connected to the in-phase output of said hybrid, respectively, said multiplier unit providing a frequency shifted in-phase spread spectrum signal having an amplitude proportional to the product of the amplitudes of the output of said medium and the amplitude of the estimated carrier generating means.

49. The matched filter of claim 47 wherein said means for downshifting includes means for downshifting said received signal with a 90°

shifted version of said estimated carrier signal, thereby providing a frequency shifted quadrature spread spectrum signal.

50. The matched filter of claim 49 wherein said means for down shifting said quadrature portion of said received signal comprises:

a hybrid with the input thereof connected to said estimated carrier generating means; and

a multiplier unit having a pair of inputs coupled to the output of said medium and connected to the quadrature output of said hybrid, respectively.

51. The matched filter of claim 43 wherein said means for downconverting includes:

means for in-phase downconverting a portion of said received signal with said in-phase carrier signal; and

an in-phase filter connected to said in-phase downconverting means for providing a signal in accordance with a relationship which is given as;

##EQU20## Where;

M_102OUT (t) is the output of said in-phase filter;

S_I1 (t) in the amplitude of a component of said received signal provided in response to said first spread baseband signal, where the carat notation indicates distortion;

d_I (t) is the amplitude of said first baseband signal;

S_Q2 (t) is the amplitude of a component of said received signal provided in response to said second spread baseband signal, where the carat notation indicates distortion; and

d_Q (t) is the amplitude of said second baseband signal.

52. The matched filter of claim 43 wherein said means for downconverting includes:

means for quadrature downconverting a portion of said received signal with said quadrature carrier signal; and

a quadrature filter connected to said quadrature demodulating means for providing a signal in accordance with a relationship which is given as;

##EQU21## Where;

M_104OUT (t) is the output of said quadrature filter;

S_I2 (t) is the amplitude of a component of said received signal provided in response to the output of said first spread sequence generator, where the carat notation indicates distortion; and

S_Q1 (t) is the amplitude of the output of a component of said received signal provided in response to the output of said second spread sequence generator, where the carat notation indicates distortion.

53. The matched filter of claim 47 wherein said approximating means is provided in response to said frequency shifted spread spectrum signal.

54. The matched filter of claim 47 wherein said approximating means includes means for providing representations of the C_n (τ

_k) terms of a relationship which is given as;

##EQU22## Where;

C_n (τ

_k) is a term of said relationship with a delay time, τ

_k ;

W_o is a difference in frequency of said carrier signal and said estimated carrier signal;

W_n is the radian frequency of a harmonic of the chip rate with an index n;

T is the chip period of said received signal;

τ

_s is a timing synchronization offset;

P(w) is the Fourier transform of a pulse provided by one of said spread spectrum sequence generators; and

H(w) is the estimate of the transfer function of said medium.

55. The matched filter of claim 47 wherein said approximating means includes a correlation unit comprising:

time delay means connected to said means for downshifting for providing said frequency shifted in-phase spread spectrum signal delayed by a known time;

first multiplier means having one input connected to said means for downshifting and another input connected to said time delay means for providing a signal having an amplitude directly related to the product of the amplitudes of said delayed frequency shifted in-phase spread spectrum signal and said frequency shifted in-phase spread spectrum signal;

second multiplier means having one input connected to said means for downshifting and another input connected to said time delay means for providing a signal having an amplitude directly related to the product of the amplitudes of said delayed frequency shifted in-phase spread spectrum signal and said frequency shifted quadrature spread spectrum signal, the outputs of said first and second multiplier means having components proportional to an in-phase part and quadrature part, respectively of the complex correlation of said received signal with said known delay time.

56. The matched filter of claim 55 wherein said approximating means includes a second down converter comprising:

reference means for generating a sinusoidal signal having a frequency equal to a harmonic of the chip rate of said received signal;

a reference hybrid with its input connected to the output of said reference means;

third multiplier means for providing a signal having an amplitude directly related to the product of the amplitudes of said in-phase part of the complex correlation of said received signal and a signal at the in-phase output of said reference hybrid;

fourth multiplier means for providing a signal having an amplitude directly related to the product of the amplitude of said quadrature part of the complex correlations of said received signal and a signal at the quadrature output of said reference hybrid;

fifth multiplier means for providing a signal having an amplitude directly related to the product of the amplitudes of said in-phase part of the complex correlation of said received signal and said signal at the quadrature output of said reference hybrid;

sixth multiplier means for providing a signal having an amplitude directly related to the product of the amplitude of said quadrature part of the complex correlation of said received signal and said signal at the in-phase output of said reference hybrid;

first adder means for providing a signal having an amplitude directly related to the sum of the amplitudes of the signals provided at the output of said third and fourth multiplier means;

second adder means for providing a signal having an amplitude directly related to the sum of the amplitudes of the signals provided at the outputs of fifth and sixth multiplier means;

first subtraction means for providing a signal directly having an amplitude directly related to the difference between the amplitudes of the signals provided at the outputs of said third and sixth multiplier means; and

second subtractor means for providing a signal directly related to the difference between the amplitudes of the signals provided at the outputs of said fourth and fifth multiplier means.

57. The matched filter of claim 43 wherein the input of said simulation means is connected to the output of said estimating means, said simulation means providing a modified first spreading signal in accordance with a relationship which is given as:

space="preserve" listing-type="equation">Ao(t)=S.sub.I1 (t)+S.sub.Q1 (t)

where;

Ao(t) is representative of the amplitude of the modified first spreading signal;

S_I1 (t) is the amplitude of an estimated component of said received signal provided in response to the output of said first spread sequence generator, where the carat notation indicates distortion by the estimated medium transfer function;

S_Q1 (t) is the amplitude of a component of said received signal provided in response to the output of said second spread sequence generator, where the carat notation indicates distortion by the estimated medium transfer function, said simulation means additionally provides a modified second spreading signal according to a relationship which is given as;

space="preserve" listing-type="equation">A.sub.90 (t)=S.sub.I2 (t)+S.sub.Q2 (t)

where;

A₉₀ (t) is representative of the amplitude of the modified second spreading signal;

S_I2 (t) is the amplitude of an estimated component of said received signal provided in response to said first spread spectrum signal where the carat notation indicates distortion by the estimated medium transfer function; and

S_Q2 (t) is the amplitude of an estimated component of said received signal provided in response to said second spread spectrum signal where the carat notation indicates distortion by the estimated medium transfer function.

58. The matched filter of claim 43 wherein said recovery means comprises:

seventh multiplier means connected to said simulation means and said downconverting means for providing a signal having an amplitude directly related to the product of the amplitude of said modified first spreading signal and the amplitude of the output of an in-phase filter that provides a signal in accordance with a relationship which is given as;

##EQU23## where;

M_102OUT (t) is the output of said in-phase filter; and

eighth multiplier means connected to said simulation means and said downconverting means for providing a signal having an amplitude directly related to the products of the amplitude of said modified first spreading signal and the amplitude of the output of a quadrature filter that provides a signal in accordance with a relationship which is given as;

##EQU24## where;

M_104OUT (t) is the output of said quadrature filter.

59. The matched filter of claim 58 wherein said recovery means additionally comprises:

ninth multiplier means connected to said simulation means and said downconverting means for providing a signal having an amplitude directly related to the product of the amplitude of said modified second spreading signal and the amplitude of the output of said in-phase filter; and

tenth multiplier means connected to said simulation means and said downconverting means for providing a signal having an amplitude directly related to the product of the amplitude of said modified second spreading signal and the amplitude of said quadrature filter.

60. The matched filter of claim 59 wherein said recovery means additionally comprises adder means for providing a signal having an amplitude directly related to the sum of the amplitude of the signals provided by said seventh and tenth multiplier means.

61. The matched filter of claim 60 wherein said recovery means additionally comprises an integrate and dump circuit connected to said adder means.

62. The matched filter of claim 59 wherein said recovery means additionally comprises subtraction means for providing a signal having an amplitude directly related to the difference between the amplitudes of signals provided by said eighth and ninth multiplier means.

63. The matched filter of claim 62 wherein said recovery means additionally comprises an integrate and dump circuit connected to said subtraction means.