Comb-assisted cyclostationary analysis

• US 10,523,329 B2
• Filed: 11/07/2017
• Issued: 12/31/2019
• Est. Priority Date: 11/07/2016
• Status: Active Grant

• Current Assignee: Regents of the University of California (University of California)
• Original Assignee: Regents of the University of California (University of California)
• Inventors: Esman, Daniel, Ataie, Vahid, Kuo, Ping Piu, Alic, Nikola, Radic, Stojan
• Primary Examiner(s): Shalaby, Mina M
• Application Number:
  US15/806,179
• Publication Number:
  US 20180131445A1
• Time in Patent Office: 784 Days
• CPC Class Codes: H04B 10/0795 Performance monitoring; Measurement of transmission parameters
• CPC Class Codes: H04B 10/5161 Combination of different modulation schemes
• CPC Class Codes: H04B 10/697 Arrangements for reducing noise and distortion

Abstract

A photonically-assisted cyclostationary analyzer eliminates the need for high-bandwidth digitization and real-time Fourier processors by using mutually-coherent frequency combs to generate a Fourier representation of the received signal in a computation-free manner.

Claims

1. A signal analyzer for analyzing an input signal to remove noise, the signal analyzer comprising:

a mutually coherent frequency comb pair comprising a first comb and a second comb, the second comb having a frequency pitch offset with respect to a frequency pitch of the first comb, each comb generating a plurality of optical tones, wherein the output of the first comb has mapped thereon the input signal, and wherein the comb pair generates a plurality of spectrally overlapping optical tone pairs, each overlapping optical tone pair comprising a single input signal copy from the first comb and a different optical tone from the second comb;

an optical demultiplexer configured for receiving the plurality of overlapping optical tone pairs and spectrally separating each overlapping optical tone pair into a plurality of sub-bands;

a detector array configured for receiving and converting each overlapping optical tone pair into an electrical frequency component of the input signal, wherein the detector array outputs a plurality of discrete Fourier transform (DFT) components corresponding to the plurality of overlapping optical tone pairs; and

a processor configured to determine a spectral correlation among the DFT components, wherein non-correlated DFT components represent noise and correlated DFT components represent signal, the processor further configured to average the correlated DFT components to generate a representation of the input signal with noise removed.

2. The signal analyzer of claim 1, wherein the input signal is in a radiofrequency (RF) domain and the CS analyzer further comprises a modulator for mapping the input signal onto the output of the first comb after comb generation.

3. The signal analyzer of claim 2, wherein the modulator comprises an electro-optical modulator.

4. The signal analyzer of claim 1, wherein the input signal is in an optical domain, and further comprising a multiplexer for combining the input signal with a seed prior to comb generation.

5. The signal analyzer of claim 1, wherein the detector array comprises a plurality of detector elements, each detector element is configured to receive at least one overlapping optical tone pair.

6. The signal analyzer of claim 5, wherein a number of detector elements of the plurality matches a number of overlapping optical tone pairs.

7. The signal analyzer of claim 5, wherein a number of detector elements of the plurality is smaller than a number of overlapping optical tone pairs.

8. The signal analyzer of claim 1, wherein the separated overlapping optical tones pairs are input into separate optical waveguides in optical communication with the detector.

18. The signal analyzer of claim 1, wherein the processor calculates spectral correlation by taking a vector outer product of the DFT component with a conjugate of itself.

9. A method for cyclostationary (CS) analysis of an input signal, the method comprising:

replicating the input signal onto a first optical frequency comb having a frequency pitch, the first optical frequency comb configured for generating a first plurality of optical tones, each optical tone of the first plurality having a replica of the input signal mapped thereon;

generating a second plurality of optical tones using a second optical frequency comb having a pitch offset relative to the first optical frequency comb, the first and second optical frequency combs being frequency-locked by a common emission seed;

splitting each replica from the first comb and at least one corresponding tone from the second comb into separate optical waveguides;

receiving the output of each optical waveguide at a coherent receiver, each optical waveguide carrying a separated replica of the input signal and at least one corresponding tone from the second comb;

converting each separated replica into an electrical sub-band of the input signal, wherein each electrical sub-band represents a discrete Fourier transform (DFT) component of the input signal;

determining a spectral correlation among the DFT components, wherein non-correlated DFT components represent noise and correlated DFT components represent signal; and

averaging correlated DFT components to generate a representation of the input signal with noise removed.

10. The method of claim 9, wherein the input signal is in a radiofrequency (RF) domain, and wherein replicating comprises mapping the input signal onto the first optical frequency comb after comb generation.

11. The method of claim 9, wherein the input signal is in an optical domain, and wherein the input signal is combined with the seed prior to generation of the first optical frequency comb.

12. The method of claim 9, wherein splitting is achieved by inputting the optical tones into an optical demultiplexer.

17. The method of claim 9, wherein calculating the spectral correlation comprises taking a vector outer product of the DFT component with a conjugate of itself.

13. A method for cyclostationary (CS) analysis of an input signal, the method comprising:

separating the input signal into a plurality of sub-bands by modulating the input signal onto a first optical frequency comb of a mutually-coherent optical comb pair, wherein the optical comb pair comprises a second optical frequency comb having a pitch offset relative to the first optical frequency comb, and wherein each sub-band corresponds to an optical tone in a spectral domain; and

detecting and converting the optical tones into electrical subcomponents of the input signal, wherein each electrical subcomponent represents a discrete Fourier transform (DFT) component of the input signal;

calculating a spectral correlation among the DFT components, wherein non-correlated DFT components represent noise and correlated DFT components represent signal; and

averaging the correlated DFT components to reconstruct the input signal.

14. The method of claim 13, wherein the input signal is in a radiofrequency (RF) domain, and wherein replicating comprises mapping the input signal onto the first optical frequency comb after comb generation.

15. The method of claim 13, wherein the input signal is in an optical domain, and wherein the input signal is combined with the seed prior to generation of the first optical frequency comb.

16. The method of claim 13, wherein calculating the spectral correlation comprises taking a vector outer product of the DFT component with a conjugate of itself.

19. A method for reconstructing an input signal, the method comprising:

spectrally decomposing the input signal into a plurality of overlapping optical tone pairs by modulating the input signal onto a first optical frequency comb of a mutually-coherent optical comb pair, wherein the mutually-coherent optical comb pair comprises a second optical frequency comb having a pitch offset relative to the first optical frequency comb;

detecting and converting the overlapping optical tone pairs into electrical subcomponents of the input signal, wherein each electrical subcomponent represents a discrete Fourier transform (DFT) component of the input signal;

calculating a spectral correlation among the DFT components, wherein non-correlated DFT components represent noise and correlated DFT components represent signal; and

averaging the correlated DFT components to reconstruct the input signal.

20. The method of claim 19, wherein detecting and converting comprises receiving the overlapping optical tone pairs in a detector array comprising a plurality of detector elements, each detector element configured to receive at least one overlapping optical tone pair.

21. The method of claim 19, further comprising, prior to detecting and converting, splitting the overlapping optical tone pairs into separate optical waveguides.

22. The method of claim 19, wherein calculating the spectral correlation comprises taking a vector outer product of the DFT component with a conjugate of itself.

23. The method of claim 19, wherein the input signal is in a radiofrequency (RF) domain, and wherein replicating comprises mapping the input signal onto the first optical frequency comb after comb generation.

24. The method of claim 19, wherein the input signal is in an optical domain, and wherein the input signal is combined with the seed prior to generation of the first optical frequency comb.