Time-Frequency Analysis With Compact Kernels for Cross-Term Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing time-frequency distributions (TFDs) for nonstationary signals face challenges in achieving high-resolution representations with minimal noise and interference, particularly due to issues with Gaussian kernels' infinite support and computational inefficiencies, as well as cross-term interference in methods like Wigner-Ville Distribution.
Innovation Solution
The Zoom algorithm combines windowed Fast Fourier Transform (FFT), spectral smoothing, convolution, and minimum smart thresholding to achieve high-resolution time-frequency analysis, using a Hamming window function and Gaussian kernel for smoothing, and adaptive thresholding to retain significant components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Gaussian kernels are used in quadratic TFDs, then time-frequency resolution is improved, but computational complexity increases and information loss occurs due to infinite support
Solution Approach 1:
The patent applies segmentation by dividing the infinite Gaussian kernel into finite segments with compact support. The kernel function is truncated to a finite interval while maintaining the essential Gaussian characteristics, thereby reducing computational complexity from O(N) to O(1) per transformation point while preserving time-frequency resolution.
Solution Approach 2:
The patent changes the parameter of kernel support from infinite to finite by introducing a compact support parameter. This parameter modification allows the kernel to maintain its smoothing and resolution properties while being computationally tractable for practical implementation in quadratic TFDs.
2Measurement precision
If Wigner-Ville Distribution is used for high-resolution representation, then time-frequency resolution is improved, but cross-term interference increases significantly
Solution Approach 1:
The patent introduces an intermediary smoothing kernel function that acts as a mediator between the signal components. This kernel convolves with the Wigner-Ville distribution to suppress cross-term interference while preserving the high resolution characteristics, effectively filtering out the harmful interference terms.
Solution Approach 2:
The patent modifies the parameter of kernel support from infinite to finite, creating a compact support kernel that reduces cross-term interference. By adjusting the support parameter, the method optimizes the balance between resolution and interference suppression, eliminating the excessive cross-terms problem of traditional WVD.
3Ease of operation
If Spectrogram method is used for time-frequency analysis, then ease of operation is improved, but time-frequency resolution tradeoff worsens
Solution Approach 1:
The patent applies dynamics by using adaptive kernel parameters that adjust based on the local characteristics of the signal. The kernel width and shape are dynamically modified to optimize time-frequency resolution for different signal segments, allowing the method to maintain ease of operation while improving resolution adaptively.
Solution Approach 2:
The patent changes the fixed window parameters of the Spectrogram method into adaptive parameters that vary with signal characteristics. By modifying the kernel parameters dynamically, the method achieves better time-frequency resolution while maintaining the computational simplicity and interpretability of the Spectrogram approach.
Data Source
AI summary
Example systems, methods, and apparatus are disclosed herein for a high-resolution time-frequency analysis for nonstationary signals. Time-frequency distributions (TFDs) are essential tools for analyzing nonstationary signals in various applications. The proposed technology, Zoom, a novel time-frequency distribution (TFD) designed to enhance the resolution and reduce cross-term interference. The performance of the proposed TFD is rigorously evaluated by studying its behavior across different smoothing values, allowing for a detailed analysis of its time-frequency localization capabilities. To quantify its effectiveness, the proposed technology computes several performance metrics, including time-frequency concentration and cross-term reduction.