Apparatus and methods for signal recovery in a digital communication system

• US 5,841,817 A
• Filed: 10/27/1995
• Issued: 11/24/1998
• Est. Priority Date: 10/27/1995
• Status: Expired due to Term

• Current Assignee: Cirrus Logic Incorporated, Commscope Emea Limited (CommScope Holding Company, Inc.), CommScope Technologies, LLC (CommScope Holding Company, Inc.)
• Original Assignee: ADC Telecommunications Incorporated (TE Connectivity Limited)
• Inventors: Kent, Mark, Krieger, Abraham
• Primary Examiner(s): Bocure, Tesfaldet
• Assistant Examiner(s): Webster, Bryan
• Application Number:
  US08/549,565
• Time in Patent Office: 1,124 Days
• Field of Search: 275/259, 275/316, 275/290, 275/340, 371/43, 375/348
• US Class Current: 375/340
• CPC Class Codes: H04L 1/0059 Convolutional codes
• CPC Class Codes: H04L 27/22 Demodulator circuits; Receiver circuits

Abstract

The present invention provides an apparatus and method for signal recovery from noisy observation in digital communication systems, utilizing a forward estimator and data smoother which improves the quality of the estimates of the transmitted signal. The present invention also provides an apparatus and method for the application of the forward estimator to the recovery of a digital data stream modulated with a CPM waveform, through the use of a receiver for GMSK demodulation. The digital data is first processed using the Forward Estimation process which calculates the probability that a symbol is in particular states. Bit decision is then applied, followed by scaling. The present invention further provides an apparatus and method for the application of the forward estimator and data smoother to the decoding of digital data which has been encoded with a Convolutional code. The digital data is first processed using the Forward Estimation procedure and then smoothed and scaled. Bit decision is then applied.

Claims

1. A method of recovering transmitted signals in a digital communications system, comprising the steps of:

(a) obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

(b) providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted signal for a respective time (p-t), where t=0 to j;

(c) storing the j vectors; and

(d) providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

2. The method of claim 1, further comprising the steps of:

(e) determining if transmission of signals has ended;

(f) if transmission has not ended, then

(f.1) storing the vector for time p and discarding the vector for time (p-j);

(f.2) obtaining additional n samples of the received signal for time (p+1);

(f.3) providing an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1); and

(f.4) providing an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

3. The method of claim 2, further comprising the step of:

recovering the transmitted signal based on the estimate vector having signal estimates obtained at time (p-j).

4. The method of claims 3, further comprising the steps of scaling the estimate vector and applying bit decision to the scaled estimate vector.

5. A method of recovering transmitted continuous phase modulated signals in a digital communications system, comprising the steps of:

(a) obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

(b) providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted continuous phase modulated signal for a respective time (p-t), where t=0 to j;

(c) storing the j vectors; and

(d) providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

6. The method of claim 5, further comprising the steps of:

(e) determining if transmission of signals has ended;

(f) if transmission has not ended, then

(f.1) storing the vector for time p and discarding the vector for time (p-j);

(f.2) obtaining additional n samples of the received signal for time (p+1);

(f.3) providing an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1); and

(f.4) providing an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

7. The method of claim 6, further comprising the step of:

recovering the transmitted signal based on the estimate vector having signal estimates for time (p-j).

8. The method of claim 7, further comprising the steps of scaling the estimate vector and applying bit decision to the scaled estimate vector.

9. The method of claim 5, wherein the step (b) comprises the steps of:

(b.1) determining phase response vector p(t_j), where

space="preserve" listing-type="equation">p(t.sub.j)= p(t.sub.0),p(t.sub.1), . . . p(t.sub.(L+2)-l-1) !

where ##EQU94## j=0, . . . (L+2)·

λ

-1, λ

is number of samples per symbol, T is symbol time and L is a predetermined integer;

(b.2) determining a first matrix G based on the phase response vector, where ##EQU95## where Gε

^2L+1, λ

(b.

     3) determining a second matrix NN, where ##EQU96## where NNε

².spsp.2L+1.sup.,2.spsp.2L+1 (b.

     4) providing a third matrix U, the third matrix being a product of the first and second matrices, where ##EQU97## (b.5) providing a probability vector comprising a first probability vector g₁ and a second probability vector g₂, where ##EQU98## where N is the number of memory states for the continuous phase modulated signal, ##EQU99##

10. The method of claim 9, wherein in the step (b.5), ##EQU100## and m/p is a modulation index.

11. The method of claim 5, wherein the step (d) comprises the steps of:

(d.1) determining a first transition vector q, where

space="preserve" listing-type="equation">q.sub.i,2·

i-1 =q.sub.i,2·

i =q.sub.1+2.sup.2L,.sub.2·

i-1 =q.sub.i+2.sup.2L,.sub.2·

i =0.5,

where i=1,2, . . . , 2^2L ;

(d.2) determining a second transition vector a, where

space="preserve" listing-type="equation">a.sub.2,1 =a.sub.N,1 =a.sub.1,N =a.sub.N-1,N =0.5

space="preserve" listing-type="equation">a.sub.i-1,i =a.sub.i+1,i =0.5, where i=2, . . . , N-1

where N is the Number of memory states for the continuous phase modulated signal; and

(d.3) providing a transition matrix QA based on the first and second transition vectors, where ##EQU101## (d.4) determining sample phasing matrices Im₁ and Im₂, where ##EQU102## and (d.5) determining a State Transition, QA_f and providing an estimate vector based on QA_f.

12. The method of claim 11, wherein in step (d.2), ##EQU103## and m/p is a modulation index.

13. A method of decoding transmitted Convolutionally encoded signals in a digital communication system, comprising the steps of:

(a) obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

(b) providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted signal for a respective time (p-t), where t=0 to j;

(c) storing the j vectors; and

(d) providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

14. The method of claim 13, further comprising the steps of:

(e) determining if transmission of signals has ended;

(f) if transmission has not ended, then

(f.1) storing the vector for time p and discarding the vector for time p-j;

(f.2) obtaining additional n samples of the received signal for time (p+1);

(f.3) providing an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1); and

(f.4) providing an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

15. The method of claim 14, further comprising the step of:

recovering the transmitted signal based on the estimate vector having signal estimates for time (p-j).

16. The method of claim 15, further comprising the steps of scaling the estimate vector and applying bit decision to the scaled estimate vector.

17. The method of claim 13, wherein step (b) comprises the steps of:

(b.1) selecting n encoding polynomials, g₁ (x), . . . , g_n (x), where x=(0, . . . , 2^mL), m/n is the code rate and mL is a size of a shift register used for data encoding;

(b.2) determining a_l, where a_l =g_l (k+i·

2^m(L-1))·

r_l, g_l being representative of the encoding polynomials g₁ (x), . . . , g_n (x), r₁ being representative of the digitized samples, and l=1, . . . , n;

(b.3) determining a function f, where f is the exponential of the summation of a_l from l=1 to l=n; and

(b.4) updating the signal estimate u, where an updated estimate u(p,k+i·

2^m(L-1)) is equal to the summation of u(p-1,k·

2m+j) taken from j=0 to j=2^m -1, multiplied by the function f.

18. The method of claim 13, wherein step (d) comprises the steps of:

(d.1) summing a plurality of probabilities of state transitions from a state at time (p-1) to a state at time p to obtain an intermediate estimate vector;

(d.2) multiplying the intermediate estimate vector with a corresponding vector obtained in step (b) component by component;

(d.3) repeating steps (d.1) and (d.2) for t=2, . . . , j, to obtain the estimate vector for time (p-j).

19. An apparatus for recovering transmitted signals in a digital communications system, comprising:

a receiver for obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

a decoder for providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted signal for a respective time (p-t), where t=0 to j;

a register for storing the j vectors; and

the decoder also providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

20. The apparatus of claim 19, wherein the decoder further determines if transmission of signals has ended, and wherein if transmission has not ended, the register further stores the vector for time p and discards the vector for time p-j, the receiver further obtains additional n samples of the received signal for time (p+1), the decoder further provides an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1), and the decoder further provides an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

21. The apparatus of claim 20, wherein the decoder further recovers the transmitted signal based on the estimate vector having signal estimates for time (p-j).

22. The apparatus of claim 21, wherein the decoder further scales the estimate vector, and applies bit decision to the scaled estimate vector.

23. An apparatus for recovering transmitted continuous phase modulated signals in a digital communications system, comprising:

a receiver for obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

a decoder for providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted continuous phase modulated signal for a respective time (p-t), where t=0 to j;

a register for storing the j vectors; and

the decoder also providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

24. The apparatus of claim 23, wherein the decoder further determines if transmission of signals has ended, and wherein if transmission has not ended, the register further stores the vector for time p and discards the vector for time p-j, the receiver further obtains additional n samples of the received signal for time (p+1), the decoder further provides an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1), and the decoder further provides an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

25. The apparatus of claim 24, wherein the decoder further recovers the transmitted signal based on the estimate vector having signal estimates for time (p-j).

26. The apparatus of claim 25, wherein the decoder further scales the estimate vector, and applies bit decision to the scaled estimate vector.

27. The apparatus of claim 23, wherein the decoder provides the j vectors by:

(b.1) determining phase response vector p(t_j), where

space="preserve" listing-type="equation">p(t.sub.j)= p(t.sub.0),p(t.sub.1), . . . p(t.sub.(L+2)-l-1) !

where ##EQU104## j=0, . . . (L+2)·

λ

-1, λ

is the number of samples per symbol, T is symbol time and L is a predetermined integer;

(b.2) determining a first matrix G based on the phase response vector, where ##EQU105## where Gε

^2L+1,λ

(b.

     3) determining a second matrix NN, where ##EQU106## where NNε

².spsp.2L+1.sup.,2.spsp.2L+1 (b.

     4) providing a third matrix U, the third matrix being a product of the first and second matrices, where ##EQU107## (b.5) providing a probability vector comprising a first probability vector g₁ and a second probability vector g₂, where ##EQU108## where N is the number of memory states for the continuous phase modulated signal, ##EQU109##

28. The apparatus of claim 27, wherein ##EQU110## and m/p is a modulation index.

29. The apparatus of claim 23, wherein the decoder provides the estimate vector by:

(d.1) determining a first transition vector q, where

space="preserve" listing-type="equation">q.sub.i,2·

i-1 =q.sub.i,2·

i =q.sub.1+2.sup.2L,.sub.2·

i-1 =q.sub.i+2.sup.2L,.sub.2·

i =0.5,

where i=1,2, . . . , 2^2L ;

(d.2) determining a second transition vector a, where

space="preserve" listing-type="equation">a.sub.2,1 =a.sub.N,1 =a.sub.1,N =a.sub.N-1,N =0.5

space="preserve" listing-type="equation">a.sub.i-1,i =a.sub.i+1,i =0.5, where i=2, . . . , N-1

where N is the Number of memory states for the continuous phase modulated signal; and

(d.3) providing a transition matrix QA based on the first and second transition vectors, where ##EQU111## (d.4) determining sample phasing matrices Im₁ and Im₂, where ##EQU112## and (d.5) determining a State Transition, QA_f and providing the estimate vector based on QA_f.

30. The apparatus of claim 29, wherein ##EQU113## where m/p is a modulation index.

31. An apparatus for decoding transmitted Convolutionally encoded signals in a digital communication system, comprising:

a receiver for obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

a decoder for providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted Convolutionally encoded signal for a respective time (p-t), where t=0 to j;

a register for storing the j vectors; and

the decoder also providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

32. The apparatus of claim 31, wherein the decoder further determines if transmission of signals has ended, and wherein if transmission has not ended, the register further stores the vector for time p and discards the vector for time p-j, the receiver further obtains additional n samples of the received signal for time (p+1), the decoder further provides an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1), and the decoder further provides an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

33. The apparatus of claim 32, wherein the decoder further recovers the transmitted signal based on the estimate vector having signal estimates for time (p-j).

34. The apparatus of claim 33, wherein the decoder further scales the estimate vector, and applies bit decision to the scaled estimate vector.

35. The apparatus of claim 31, wherein the decoder provides the j vectors by:

(b.1) selecting n encoding polynomials, g₁ (x), . . . . , g_n (x), where x=(0, . . . , 2^mL), m/n is the code rate and mL is a size of a shift register used for data encoding;

(b.2) determining a_l, where a_l =g_l (k+i·

2^m(L-1))·

r_l, g_l being representative of the encoding polynomials g₁ (x), . . . , g_n (x), r_l being representative of the digitized samples, and l=1, . . . , n;

(b.3) determining a function f, where f is the exponential of the summation of a_l from l=1 to l=n; and

(b.4) updating the signal estimate u, where an updated estimate u(p,k+i·

2^m(L-1)) is equal to the summation of u(p-1,k·

2m+j) taken from j=0 to j=2^m -1, multiplied by the function f.

36. The apparatus of claim 31, wherein the decoder provides the estimate vector by:

(d.1) summing a plurality of probabilities of state transitions from a state at time (p-1) to a state at time p to obtain an intermediate estimate vector;

(d.2) multiplying the intermediate estimate vector with a corresponding vector obtained in step (b) component by component;

(d.3) repeating steps (d.1) and (d.2) for t=2, . . . , j, to obtain the estimate vector for time (p-j).

37. Computer-executable process steps for recovering transmitted signals in a digital communications system, the process steps including:

(a) obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

(b) providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted signal for a respective time (p-t), where t=0 to j;

(c) storing the j vectors; and

(d) providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

38. The computer-executable process steps of claim 37, further including the steps of:

(e) determining if transmission of signals has ended;

(f) if transmission has not ended, then

(f.1) storing the vector for time p and discarding the vector for time p-j;

(f.2) obtaining additional n samples of the received signal for time (p+1);

(f.3) providing an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1); and

(f.4) providing an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

39. The computer-executable process steps of claim 38, further including the step of:

recovering the transmitted signal based on the estimate vector having signal estimates for time (p-j).

40. The computer-executable process steps of claim 39, further including the steps of:

scaling the estimate vector; and

applying bit decision to the scaled estimate vector.

41. Computer-executable process steps for recovering transmitted continuous phase modulated signals in a digital communications system, including the steps of:

(a) obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

(b) providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted continuous phase modulated signal for a respective time (p-t), where t=0 to j;

(c) storing the j vectors; and

(d) providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

42. The computer-executable process steps of claim 41, further including the steps of:

(e) determining if transmission of signals has ended;

(f) if transmission has not ended, then

(f.1) storing the vector for time p and discarding the vector for time p-j;

(f.2) obtaining additional n samples of the received signal for time (p+1);

(f.3) providing an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1); and

(f.4) providing an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

43. The computer-executable process steps of claim 42, further including the step of:

recovering the transmitted signal based on the estimate vector having signal estimates for time (p-j).

44. The computer-executable process steps of claim 43, further including the steps of:

scaling the estimate vector; and

applying bit decision to the scaled estimate vector.

45. The computer-executable process steps of claim 41, wherein the step (b) comprises the steps of:

(b.1) determining phase response vector p(t_j), where

space="preserve" listing-type="equation">p(t.sub.j)= p(t.sub.0),p(t.sub.1), . . . p(t.sub.(L+2)-l-1) !

where ##EQU114## j=0, . . . (L+2)·

λ

-1, λ

is the number of samples per symbol, T is symbol time and L is a predetermined integer;

(b.2) determining a first matrix G based on the phase response vector, where ##EQU115## where Gε

^2L+1,λ

(b.

     3) determining a second matrix NN, where ##EQU116## where NNε

².spsp.2L+1.sup.,2.spsp.2L+1 (b.

     4) providing a third matrix U, the third matrix being a product of the first and second matrices, where ##EQU117## (b.5) providing a probability vector comprising a first probability vector g₁ and a second probability vector g₂, where ##EQU118## N is the number of memory states of the continuous phase modulated signal, ##EQU119##

46. The computer-executable process steps of claim 41, wherein in the step (b.5), ##EQU120## where m/p is a modulation index.

47. The computer-executable process steps of claim 41, wherein the step (d) comprises the steps of:

(d.1) determining a first transition vector q, where

space="preserve" listing-type="equation">q.sub.i,2·

i-1 =q.sub.i,2·

i =q.sub.i+2.sup.2L,.sub.2·

i-1 =q.sub.i+2.sup.2L,.sub.2·

i =0.5,

where i=1,2, . . . ,2^2L ;

(d.2) determining a second transition vector a, where

space="preserve" listing-type="equation">a.sub.2,1 =a.sub.N,1 =a.sub.1,N =a.sub.N-1,N =0.5

space="preserve" listing-type="equation">a.sub.i-1,i =a.sub.i+1,i =0.5, where i=2, . . . , N-1

where N is the Number of memory states of the continuous phase modulated signal; and

(d.3) providing a transition matrix QA based on the first and second transition vectors, where ##EQU121## (d.4) determining sample phasing matrices Im₁ and Im₂, where ##EQU122## and (d.5) determining a State Transition, QA_f and providing the estimate vector based on QA_f.

48. The computer-executable process steps of claim 47, wherein in step (d.2), ##EQU123## where m/p is a modulation index.

49. Computer-executable process steps of decoding transmitted Convolutionally encoded signals in a digital communication system, including the steps of:

(a) obtaining n samples of a received signal for a respective time p, p-1, p-2, . . . , p-j and digitizing the n samples to provide (j+1) sets of n digitized samples;

(b) providing (j+1) vectors based on the (j+1) sets of n digitized samples, each vector having signal estimates of a transmitted Convolutionally encoded signal for a respective time (p-t), where t=0 to j;

(c) storing the j vectors; and

(d) providing an estimate vector having signal estimates of the transmitted signal for time (p-j), based on the (j+1) vectors.

50. The computer-executable process steps of claim 49, further including the steps of:

(e) determining if transmission of signals has ended;

(f) if transmission has not ended, then

(f.1) storing the vector for time p and discarding the vector for time p-j;

(f.2) obtaining additional n samples of the received signal for time (p+1);

(f.3) providing an additional vector based on the additional n samples, the vector having signal estimates of the transmitted signal for time (p+1); and

(f.4) providing an estimate vector having signal estimates of the transmitted signal for time (p+1-j), based on the additional vector and the immediately preceding j vectors.

51. The computer-executable process steps of claim 50, further including the step of:

recovering the transmitted signal based on the estimate vector having signal estimates for time (p-j).

52. The computer-executable process steps of claim 51, further including the steps of:

scaling the estimate vector; and

applying bit decision to t he scaled estimate vector.

53. The computer-executable process steps of claim 49, wherein step (b) includes the steps of:

(b.1) selecting n encoding polynomials, g₁ (x), . . . , g_n (x), where x=(0, . . . ,2^mL), m/n is the code rate and mL is a size of a shift register used for data encoding;

(b.2) determining a_l, where a_l =g_l (k+i·

2^m(L-1))·

r_l, g_l being representative of the encoding polynomials g₁ (x), . . . , g_n (x), r₁ being representative of the digitized samples, and l=1, . . . ,n;

(b.3) determining a function f, where f is the exponential of the summation of a_l from l=1 to l=n; and

(b.4) updating the signal estimate u, where an updated estimate u(p,k+i·

2^m(L-1)) is equal to the summation of u(p-1,k·

2m+j) taken from j=0 to j=2^m -1, multiplied by the function f.

54. The computer-executable process steps of claim 53, wherein step (d) includes the steps of:

(d.1) summing a plurality of probabilities of state transitions from a state at time (p-1) to a state at time p to obtain an intermediate estimate vector;

(d.2) multiplying the intermediate estimate vector with a corresponding vector obtained in step (b) component by component;

(d.3) repeating steps (d.1) and (d.2) for t=2, . . . , j, to obtain the estimate vector for time (p-j).