Method for data compression and time-bandwidth product engineering

Patent metadata

Field	Value
Publication number	US-2016006453-A1
Application number	US-201514749626-A
Country	US
Kind code	A1
Filing date	Jun 24, 2015
Priority date	Dec 27, 2012
Publication date	Jan 7, 2016
Grant date	—

Abstract

A method and apparatus for compression of digital and analog data utilizing an anamorphic spectrum transformation to warp an analog or digital signal to provide time-bandwidth compression after sampling is performed on the warped signal. The anamorphic spectrum transformation performs reallocating samples to assign more samples to fine features with high frequency waveform portions and fewer samples to coarse features with low frequency portions of the analog temporal signal, resulting in shorter record length and fewer number of samples.

First claim

What is claimed is:

1 . A method of capturing a temporal waveform, comprising:
receiving a temporal waveform at a transformation stage; and performing a transformation at said transformation stage which imparts a nonlinear warp onto the frequency spectrum of the waveform to output a compressed temporal waveform.

2 . The method recited in claim 1 , wherein said transformation stage comprises a warped/anamorphic phase operation.

3 . The method recited in claim 1 , wherein said transformation stage comprises a filter with non-quadratic phase response.

4 . The method recited in claim 1 , wherein the temporal waveform is nonuniformly transformed during said nonlinear warping so that a uniform and finite sampling rate matches features of the temporal waveform.

5 . The method recited in claim 4 , wherein during said nonlinear warping, spectrum in a Fourier domain of the temporal waveform is warped to a spectrum probability density function (SPDF).

6 . The method recited in claim 1 , wherein said method is configured for increasing the bandwidth of data converters without reducing record length or requiring a higher number of samples.

7 . The method recited in claim 1 , wherein said transformation stage is utilized for imparting said nonlinear warp to said temporal waveform prior to receipt on a conventional data converter to provide adaptive temporal sampling in which because of the transformation, sampling rate adapts to the temporal waveform so that fine temporal features are sampled at a higher rate than the coarse temporal features.

8 . The method recited in claim 7 , wherein said data converter comprises capturing of the transformed temporal waveform with an analog to digital converter (ADC).

9 . The method recited in claim 1 , further comprising a decompression of said compressed temporal waveform.

10 . The method recited in claim 9 , wherein said decompression performs the inverse of said transformation on a compressed temporal waveform received from a conventional data converter.

11 . The method recited in claim 10 , wherein said conventional data converter comprises a digital to analog converter (DAC).

12 . The method recited in claim 1 , wherein said transformation stage is performed in the analog domain prior to sampling, and does not suffer from either quantization noise or the limits of a finite number of samples and quantization noise of a data converter.

13 . The method recited in claim 1 , wherein said transformation is performed in one or more domains selected from a group of domains consisting of analog, digital, optical, and electrical/radio/microwave frequency.

14 . The method recited in claim 1 , wherein said transformation is performed on digital data with all-digital signal processing.

15 . The method recited in claim 1 , wherein said nonlinear warping is performed using a filter having a specific transfer function; wherein unless this specific transfer function is known, an eavesdropper would have difficulty recovering the temporal waveform.

16 . The method recited in claim 1 , wherein said method enhances resolution of spectroscopy, or decreases required number of samples without detracting from equivalent system resolution.

17 . The method recited in claim 1 , wherein said transformation exploits a natural frequency-to-time relation inherent in temporal dispersion, and does not rely on a-priori knowledge of the temporal waveform.

18 . A method for increasing temporal resolution of data converters, comprising:
receiving a temporal waveform at a transformation stage; performing a transformation at said transformation stage utilizing frequency-selective mapping of the spectrum, with nonlinear warping, whereby mapping scale has a larger stretch factor on fine temporal features, subjecting said temporal waveform to an injective group delay to produce a compressed temporal waveform; and receiving said compressed temporal waveform at a data converter wherein it is sampled with a uniform sample pattern.

19 . The method recited in claim 18 , wherein said injective group delay corresponds to a phase response that depends on a combination of even-order powers of frequency with weighting factors.

20 . The method recited in claim 18 , wherein said transformation is performed in one or more domains selected from a group of domains consisting of analog, digital, optical, and electrical/radio/microwave frequency.

21 . The method recited in claim 18 , wherein said transformation reduces number of samples required to obtain a given temporal bandwidth.

22 . The method recited in claim 18 , wherein said transformation is performed on digital data with all-digital signal processing.

23 . The method recited in claim 18 , wherein said nonlinear warping is performed using a filter having a specific transfer function, whereby unless this specific transfer function is known, an eavesdropper would have difficulty recovering the temporal waveform.

24 . The method recited in claim 18 , wherein said method enhances resolution of spectroscopy, or decreases number of samples required, without adversely impacting equivalent system resolution.

25 . The method recited in claim 18 , wherein said transformation exploits a natural frequency-to-time relation inherent in temporal dispersion, and does not rely on a-priori knowledge of the temporal waveform.

Assignees

• Univ California

Inventors

• Jalali Bahram
• Asghari Mohammad H

Classifications

• H03M7/3059

  Digital compression and data reduction techniques where the original information is represented by a subset or similar information, e.g. lossy compression · CPC title

• H03M5/22 Primary

  Conversion to or from representation by sinusoidal signals · CPC title