Device to process sample using a time-windowed transform function to generate spectral data and to use combined magnitude and phase spectrograms

1. A device for signal processing comprising:

a memory configured to store instructions; and

one or more processors configured to execute the instructions to:

receive samples of a signal;

process the samples using a time-windowed transform function to generate spectral data corresponding to each time window;

generate first spectrogram data based on magnitudes of the spectral data;

generate, from the spectral data, phase data that is corrected for phase rotation due to frequency offset from baseband;

generate phase difference data based on the phase data;

perform an unwrapping operation of the phase difference data to generate unwrapped difference data;

rectify the unwrapped difference data to generate second spectrogram data;

combine the first spectrogram data and the second spectrogram data to generate a combined spectrogram; and

process the combined spectrogram to generate an output that identifies one or more signal types associated with the signal, identifies one or more emitters associated with the signal, or both.

2. The device of claim 1 , wherein the one or more processors are configured to multiply values of the first spectrogram data and values of the second spectrogram data to generate the combined spectrogram.

3. The device of claim 1 , wherein the one or more processors are configured to generate multiple combined spectrograms for the signal using different time window sizes of the transform function.

4. The device of claim 3 , wherein the one or more processors are configured to format the multiple combined spectrograms as multi-channel image data for processing at an image processor.

5. The device of claim 4 , wherein each particular bin of one channel of the multi-channel image data corresponds to a same time and frequency block as the corresponding bins of the other channels of the multi-channel image data.

6. The device of claim 1 , wherein the combined spectrogram provides a high-fidelity image of signal information.

7. The device of claim 1 , wherein the one or more processors are further configured to send the combined spectrogram to a display device for visual display.

8. The device of claim 1 , wherein the one or more processors are further configured to input the combined spectrogram into a neural network configured to process image data.

9. The device of claim 8 , wherein the combined spectrogram enables detection, by the neural network, of one or more waveform signatures, and wherein the detection of one or more waveform signatures includes preamble detection.

10. The device of claim 9 , wherein the combined spectrogram enables identification, by the neural network, of a particular waveform instance based on the one or more waveform signatures.

11. The device of claim 8 , wherein the combined spectrogram enables detection, by the neural network, of a presence of multiple emitting sources.

12. The device of claim 11 , wherein the one or more processors are configured to identify specific emitting sources of the multiple emitting sources based on detecting, within the combined spectrogram, characteristics unique to the specific emitting sources.

13. A method of signal processing comprising:

receiving samples of a signal;

processing the samples using a time-windowed transform function to generate spectral data corresponding to each time window;

generating first spectrogram data based on magnitudes of the spectral data;

generating, from the spectral data, phase data that is corrected for phase rotation due to frequency offset from baseband;

generating phase difference data based on the phase data;

performing an unwrapping operation of the phase difference data to generate unwrapped difference data;

rectifying the unwrapped difference data for generating second spectrogram data based on phase differences of the spectral data;

combining the first spectrogram data and the second spectrogram data to generate a combined spectrogram; and

processing the combined spectrogram to generate an output that identifies one or more signal types associated with the signal, identifies one or more emitters associated with the signal, or both.

14. The method of claim 13 , further comprising generating multiple combined spectrograms for the signal using different time window sizes of the transform function.

15. The method of claim 14 , further comprising formatting the multiple combined spectrograms as multi-channel image data for processing at an image processor.

16. The method of claim 13 , wherein the combined spectrogram enables detection, by a neural network, of one or more waveform signatures, and wherein the detection of one or more waveform signatures includes preamble detection.

17. The method of claim 13 , further comprising multiplying values of the first spectrogram data and values of the second spectrogram data to generate the combined spectrogram.

18. The method of claim 13 , further comprising inputting the combined spectrogram into a neural network configured to process image data, wherein the combined spectrogram enables detecting, by the neural network, of one or more waveform signatures, and wherein the detecting of one or more waveform signatures includes preamble detection.

19. A computer-readable storage device storing instructions that, when executed by one or more processors, cause the one or more processors to:

receive samples of a signal;

process the samples using a time-windowed transform function to generate spectral data corresponding to each time window;

generate first spectrogram data based on magnitudes of the spectral data;

generate, from the spectral data, phase data that is corrected for phase rotation due to frequency offset from baseband;

generate phase difference data based on the phase data;

perform an unwrapping operation of the phase difference data to generate unwrapped difference data;

rectify the unwrapped difference data to generate second spectrogram data based on phase differences of the spectral data;

combine the first spectrogram data and the second spectrogram data to generate a combined spectrogram; and

process the combined spectrogram to generate an output that identifies one or more signal types associated with the signal, identifies one or more emitters associated with the signal, or both.

20. The computer-readable storage device of claim 19 , wherein the instructions, when executed by the one or more processors, further cause the one or more processors to generate multiple combined spectrograms for the signal using different time window sizes of the transform function.