Direct frequency selection and down-conversion for digital receivers

• US 5,982,823 A
• Filed: 03/17/1998
• Issued: 11/09/1999
• Est. Priority Date: 03/17/1998
• Status: Expired due to Term

• Current Assignee: Northrop Grumman Systems Corporation (Northrop Grumman Corporation)
• Original Assignee: Northrop Grumman Corporation
• Inventors: Jacklin, William Edward
• Primary Examiner(s): Pham, Chi H.
• Assistant Examiner(s): Bayard, Emmanuel
• Application Number:
  US09/044,202
• Time in Patent Office: 602 Days
• Field of Search: 375/316, 375/344, 375/261, 375/267, 455/182.1, 455/182.3, 455/192.1, 455/192.3, 455/130, 364/724.1, 364/723
• US Class Current: 375/344
• CPC Class Codes: H03D 3/007 by converting the oscillations into two quadrature related signals H03D3/245 takes precedence
• CPC Class Codes: H04B 1/30 for homodyne or synchrodyne receivers demodulator circuits H03D1/22

Abstract

A method and a coarse tuning circuit for performing direct frequency selection and down-conversion on a received analog bandpass signal for a digital receiver having C channels are disclosed. Each of the C channels has a bandwidth B. The analog bandpass signal in a channel of interest has a center frequency f_c and a two-sided bandwidth not greater than B. The method comprises the following steps: (a) sampling the bandpass signal at a sampling frequency f_s greater than twice the product of B and C, to produce a digital sampled signal which has an intermediate frequency much closer to the baseband frequency of zero kilohertz than the sampling frequency f_s ; (b) computing, from f_c and f_s, a non-negative integer value m to determine the selection of a channel from the C channels of the digital receiver; and (c) forming directly from the digital sampled signal two digital signals I₁ and Q₁ such that they correspond to a digital in-phase component and a quadrature component, respectively, of a complex signal which is approximately centered at a frequency Δf proximate of the baseband frequency, and which has a frequency spectrum approximately equal to that of the analog bandpass signal.

Claims

1. A method for performing direct frequency selection and down-conversion on an analog bandpass signal for a digital receiver having C channels, each of the C channels having a bandwidth B, frequency spectrum of the analog bandpass signal having a center frequency f_c and a two-sided bandwidth not greater than B in one of the C channels, the method comprising:

(a) sampling the analog bandpass signal at a sampling frequency f_s, f_s being greater than twice the product of B and C, to produce a digital sampled signal, the digital sampled signal having an intermediate frequency proximate of a baseband frequency, the intermediate frequency being closer to the baseband frequency than the sampling frequency f_s ;

(b) computing, from f_c and f_s, a non-negative integer value m, said value m determining selection of a channel from the C channels of the digital receiver;

(c) forming from said digital sampled signal {r(kT_s), with T_s =1/f_s and k=0, 1, . . . } a sequence of even samples {r(kT_s), with T_s =1/f_s and k=0, 2, 4, . . . } and a sequence of odd samples {r(kT_s), with T_s =1/f_s and k=1, 3, 5, . . . };

(d) generating from the sequences of even samples and odd samples a sequence I and a sequence Q such that the sequences I and Q are approximately equal to an in-phase component and a quadrature component, respectively, of a complex signal centered near the baseband frequency having a frequency spectrum approximately equal to the analog bandpass signal frequency spectrum; and

(e) forming first digital signals I₁ and Q₁ by interpolating at least one digital value between every two consecutive samples of the sequences I and Q, respectively resulting in each of the first digital signals I₁ and Q₁ being lowpass filtered and having a sampling rate greater than that of each of the sequences I and Q.

2. The method for performing direct frequency selection and down-conversion as recited in claim 1 wherein the sequences I and Q are generated according to the following formula:

space="preserve" listing-type="equation">I={r(0), -r(2T.sub.s), r(4T.sub.s), -r(6T.sub.s), . . . },

space="preserve" listing-type="equation">if m is even, Q={-r(T.sub.s), r(3T.sub.s), -r(5T.sub.s), r(7T.sub.s), . . . },

space="preserve" listing-type="equation">if m is odd, Q={r(T.sub.s), -r(3T.sub.s), r(5T.sub.s), -r(7T.sub.s), . . . },

where m is the integer rounded-off value of the expression (4f_c -f_s)/2f_s.

3. The method for performing direct frequency selection and down-conversion as recited in claim 1 further comprising the steps of:

(1) shifting in complex frequency the first digital signals I₁ and Q₁ by an amount equal to Δ

f to form second digital signals I₂ and Q₂ ; and

(2) outputting the second digital signals I₂ and Q₂ which correspond to an in-phase component and a quadrature component, respectively, of a complex baseband frequency signal, the complex baseband frequency signal having a frequency spectrum, centered at the baseband frequency, approximately equal to the analog bandpass signal frequency spectrum centered at the frequency f_c.

4. The method as recited in claim 1 wherein the sampling frequency f_s is equal to 4f_c /(2m+1).

5. The method as recited in claim 1 wherein the sampling frequency f_s is not equal to 4f_c /(2m+1).

6. The method as recited in claim 1 wherein the sampling frequency f_s is such that the intermediate frequency of the digital sampled signal differs from the baseband frequency by an amount equal to (f_s /4)+Δ

f.

7. The method as recited in claim 1 wherein the frequency Δ

f is equal to (2m+1) (f_s /4)-f_c, where m is the integer rounded-off value of the expression (4f_c -f_s)/2f_s.

8. The method as recited in claim 3 wherein the second digital signals I₂ and Q₂ are formed from the first digital signals I₁ and Q₁ according to the following formula:

space="preserve" listing-type="equation">I.sub.2 (kT.sub.s)=I.sub.1 (kT.sub.s)cos(2π

Δ

fkT.sub.s)-Q.sub.1 (kT.sub.s)sin(2π

Δ

fkT.sub.s),

space="preserve" listing-type="equation">Q.sub.2 (kT.sub.s)=Q.sub.1 (kT.sub.s)cos(2π

Δ

fkT.sub.s)+I.sub.1 (kT.sub.s)sin(2π

Δ

fkT.sub.s),

space="preserve" listing-type="equation">for k=0, 1, 2, . . . ,

where T_s =1/f_s.

9. The method as recited in claim 1 wherein the baseband frequency is equal to zero kilohertz.

10. A coarse tuning circuit for a digital receiver having C channels, each of the C channels having a bandwidth B, the coarse tuning circuit down-converting to a frequency near a baseband frequency a digital sampled signal formed by sampling an analog bandpass signal at a sampling frequency f_s, frequency spectrum of the analog bandpass signal having a center frequency f_c and a two-sided bandwidth not greater than B in one of the C channels, the sampling frequency f_s being greater than twice the product of B and C, the coarse tuning circuit comprising:

(a) a selecting circuit for computing, from f_c and f_s, a non-negative integer value m, said value m determining selection of a channel from the C channels of the digital receiver;

(b) a switching circuit having an input port to receive said digital sampled signal {r(kT_s), with T_s =1/f_s and k=0, 1, . . . }, and an output port to produce a sequence of even samples and a sequence of odd samples, said sequences being formed from the received digital sampled signal;

(c) a sign correction circuit, in electrical communication with the selecting circuit and with the switching circuit to receive the non-negative integer value m and the sequences of even samples and odd samples, for generating a sequence I and a sequence Q such that the sequences I and Q are approximately equal to an in-phase component and a quadrature component, respectively, of a complex signal centered near the baseband frequency having a frequency spectrum approximately equal to the analog bandpass signal frequency spectrum; and

(d) an interpolation filter circuit, in electrical communication with the sign correction circuit to receive the sequences I and Q, for forming output digital signals I₁ and Q₁ by interpolating at least one digital value between every two consecutive samples of the sequences I and Q, respectively, resulting in each of the output digital signals I₁ and Q₁ being lowpass filtered and having a sampling rate greater than that of each of the sequences I and Q.

11. The coarse tuning circuit as recited in claim 10 wherein the non-negative integer value m computed by the selecting circuit is equal to the integer rounded-off value of the expression (4f_c -f_s)/2f_s.

12. The coarse tuning circuit as recited in claim 10 wherein the sign correction circuit generates the sequences I and Q according to the following formula:

space="preserve" listing-type="equation">I={r(0), -r(2T.sub.s), r(4T.sub.s), -r(6T.sub.s), . . . },

space="preserve" listing-type="equation">if m is even, Q={-r(T.sub.s), r(3T.sub.s), -r(5T.sub.s), r(7T.sub.s), . . . },

space="preserve" listing-type="equation">if m is odd, Q={r(T.sub.s), -r(3T.sub.s), r(5T.sub.s), -r(7T.sub.s), . . . }.

13. The coarse tuning circuit as recited in claim 12 wherein the sign correction circuit translates an image signal centered at a frequency proximate of f_s /4 toward the baseband frequency by an amount equal to f_s /4, said image signal being included in the digital sampled signal and having a frequency spectrum approximately equal to the analog bandpass signal frequency spectrum.

14. The coarse tuning circuit as recited in claim 12 wherein the sampling frequency f_s is equal to 4f_c /(2m+1).

15. The coarse tuning circuit as recited in claim 12 wherein the sampling frequency f_s is not related to the frequency f_c.

16. The coarse tuning circuit as recited in claim 10 wherein the baseband frequency is equal to zero kilohertz.

17. A system for digitally performing direct frequency selection and down-conversion on an analog bandpass signal for a digital receiver having C channels, each of the C channels having a bandwidth B, frequency spectrum of the analog bandpass signal having a center frequency f_c and a two-sided bandwidth not greater than B in one of the C channels, the system comprising:

(a) an analog-to-digital (A/D) converter for sampling the analog bandpass signal at a sampling frequency f_s greater than twice the product of B and C, and for outputting a digital sampled signal, the digital sampled signal having an intermediate frequency closer to a baseband frequency than f_s ;

(b) a coarse tuning circuit, in electrical communication with the A/D converter, for shifting the intermediate frequency of the digital sampled signal toward the baseband frequency by generating two first digital signals I₁ and Q₁, the shifted intermediate frequency differing from the baseband frequency by an amount Δ

f; and

(c) a fine tuning circuit, in electrical communication with the coarse tuning circuit, for shifting in complex frequency toward the baseband frequency the first digital signals I₁ and Q₁ by the amount Δ

f to form second digital signals I₂ and Q₂ the second digital signals I₂ and Q₂ being approximately equal to an in-phase component and a quadrature component, respectively, of a digital baseband frequency signal, the digital baseband frequency signal having a frequency spectrum, centered at the baseband frequency, approximately equal to the analog bandpass signal frequency spectrum centered at frequency f_c

(d) wherein the coarse tuning circuit comprises;

(i) a selecting circuit for computing, from f_c and f_s, a non-negative integer value m, said value m determining selection of a channel from the C channels of the digital receiver;

(ii) a switching circuit having an input port to receive said digital sampled signal {r(kT_s), with T_s =1f_s and k=0, 1, . . . }, and an output port to produce a sequence of even samples and a sequence of odd samples, said sequences being formed from the received digital sampled signal;

(iii) a sign correction circuit, in electrical communication with the selecting circuit and with the switching circuit to receive the non-negative integer value m and the sequences of even samples and odd samples, for generating a sequence I and a sequence Q such that the sequences I and Q are approximately equal to an in-phase component and a quadrature component, respectively, of a complex signal centered near the baseband frequency having a frequency spectrum approximately equal to the analog bandpass signal frequency spectrum; and

(iv) an interpolation filter circuit, in electrical communication with the sign correction circuit to receive the sequences I and Q, for forming first digital signals I₁ and Q₁ by interpolating at least one digital value between every two consecutive samples of the sequences I and Q, respectively, resulting in each of the first digital signals I₁ and Q₁ being lowpass filtered and having a sampling rate greater than that of each of the sequences I and Q.

18. The system as recited in claim 17 wherein the sampling frequency f_s is equal to 4f_c /(2m+1).

19. The system as recited in claim 17 wherein the sampling frequency f_s is not equal to 4f_c (2m+1).

20. The system as recited in claim 17 wherein the sampling frequency f_s is such that the intermediate frequency of the digital sampled signal differs from the baseband frequency by an amount equal to (f_s /4)+Δ

f.

21. The system as recited in claim 17 wherein the non-negative integer value m computed by the selecting circuit is equal to the integer rounded-off value of the expression (4f_c -f_s)/2f_s.

22. The system as recited in claim 21 wherein the sign correction circuit generates the sequences I and Q according to the following formula:

space="preserve" listing-type="equation">I={r(0), -r(2T.sub.s), r(4T.sub.s), -r(6T.sub.s), . . . },

space="preserve" listing-type="equation">if m is even, Q={-r(T.sub.s), r(3T.sub.s), -r(5T.sub.s), r(7T.sub.s), . . . },

space="preserve" listing-type="equation">if m is odd, Q={r(T.sub.s), -r(3T.sub.s), r(5T.sub.s), -r(7T.sub.s), . . . }.

23. The system as recited in claim 18 wherein the sign correction circuit translates an image signal centered at a frequency proximate of f_s /4 toward the baseband frequency by an amount equal to f_s /4, said image signal being included in the digital sampled signal and having a frequency spectrum approximately equal to the analog bandpass signal frequency spectrum.

24. The system as recited in claim 17 wherein the coarse tuning circuit computes the value of Δ

f according to the following formula;

space="preserve" listing-type="equation">Δ

f=(2m+1) (f.sub.s /4)-f.sub.c

where m is the integer rounded-off value of the expression (4f_c -f_s)/2f_s.

25. The system as recited in claim 17 wherein the fine tuning circuit forms the second digital signals I₂ and Q₂ from the first digital signals I₁ and Q₁ according to the following formula:

space="preserve" listing-type="equation">I.sub.2 (kT.sub.s)=I.sub.1 (kT.sub.s)cos(2π

Δ

fkT.sub.s)-Q.sub.1 (kT.sub.s)sin(2π

Δ

fkT.sub.s),

space="preserve" listing-type="equation">Q.sub.2 (kT.sub.s)=Q.sub.1 (kT.sub.s)cos(2π

Δ

fkT.sub.s)+I.sub.1 (kT.sub.s)sin(2π

Δ

fkT.sub.s),

space="preserve" listing-type="equation">for k=0, 1, 2, . . . , where T.sub.s =1/f.sub.s.

26. The system as recited in claim 17 further comprising a lowpass filter circuit for receiving and filtering the second digital signals I₂ and Q₂.

27. The system as recited in claim 17 wherein the baseband frequency is equal to zero kilohertz.