[Paper Review] A New Look at an Old Tool-the Cumulative Spectral Power of Fast-Fourier Transform Analysis

Sheng-Chiang Lee, Randall D. Peters|ArXiv.org|Jan 23, 2009

Heart Rate Variability and Autonomic Control2 references5 citations

TL;DR

This paper introduces Cumulative Spectral Power (CSP) as a novel approach to enhance Fast Fourier Transform (FFT) analysis for time-varying, noisy signals. By cumulatively summing squared spectral amplitudes up to each frequency, CSP provides a more robust representation than traditional Power Spectral Density (PSD), offering improved sensitivity to transient features and low-amplitude components in noisy data across diverse applications.

ABSTRACT

As an old and widely used tool, it is still possible to find new insights and applications from Fast Fourier Transform (FFT)-based analyses. The FFT is frequently used to generate the Power Spectral Density (PSD) function, by squaring the spectral components that have been corrected for influence from the instrument that generated the data. Although better than a raw-data spectrum, by removing influence of the instrument transfer function, the PSD is still of limited value for time varying signals with noise, due to the very nature of the Fourier transform. The authors present here another way to treat the FFT data, namely the Cumulative Spectral Power (CSP), as a promising means to overcome some of these limitations. As will be seen from the examples provided, the CSP holds promise in a variety of different fields.

Motivation & Objective

• To address the limitations of Power Spectral Density (PSD) in analyzing time-varying, noisy signals using FFT.
• To develop a more sensitive and robust spectral representation that captures transient and low-amplitude features often obscured in standard FFT-based methods.
• To demonstrate the utility of Cumulative Spectral Power (CSP) across diverse scientific and engineering applications where signal clarity is critical.
• To provide an alternative to PSD that better preserves information in non-stationary and noisy data by cumulatively aggregating spectral energy.

Proposed method

• The authors compute the cumulative sum of squared FFT spectral components up to each frequency bin, forming the Cumulative Spectral Power (CSP) function.
• CSP is derived from the FFT output by sequentially summing the squared magnitudes of frequency components, starting from the lowest frequency.
• The method corrects for instrument transfer function effects, similar to PSD, but emphasizes energy accumulation rather than spectral density at individual frequencies.
• The approach is applied to real-world data examples to illustrate its ability to reveal subtle signal features not easily detectable via standard PSD.
• CSP is compared visually and qualitatively with PSD to highlight its enhanced sensitivity to low-energy and transient components.
• The technique is implemented using standard FFT libraries, with post-processing to compute the cumulative sum of squared amplitudes.

Experimental results

Research questions

• RQ1 Can cumulative aggregation of spectral power improve the detection of weak or transient features in noisy, time-varying signals compared to standard PSD?
• RQ2 How does CSP perform in identifying subtle spectral components that are obscured in traditional FFT-based spectral density analysis?
• RQ3 In what types of real-world data does CSP provide a more informative representation than conventional spectral power analysis?
• RQ4 Does the cumulative nature of CSP enhance the interpretability of spectral data in non-stationary or high-noise environments?

Key findings

• CSP provides a more sensitive representation of spectral energy distribution than PSD, particularly for low-amplitude and transient signal components.
• The cumulative summation process enhances visibility of weak spectral features that are masked in standard PSD plots due to noise or dynamic range limitations.
• Examples from the paper demonstrate that CSP reveals structural patterns in time-varying signals that are not apparent in conventional spectral density representations.
• The method shows promise across diverse fields, including physics, engineering, and data analysis, where signal clarity and feature detection are critical.
• CSP effectively reduces the impact of spectral leakage and noise by emphasizing energy accumulation over frequency bands rather than individual frequency bins.
• The technique maintains compatibility with standard FFT processing pipelines, enabling straightforward integration into existing spectral analysis workflows.