A Review of Signal Detection Using the Bispectrum with Applications in Underwater Acoustics
Abstract
This paper reviews attempts at signal detection in Gaussian noise using a higher order statistical (Higher Order Spectra (HOS) or polyspectra) technique. Examples comparing power spectral and bispectral analysis include the following topics: the identification of signals generated by a system of coupled nonlinear differential equations, radar backscatter processing and target identification, and a statistical treatment of the detection of narrowband harmonic components resulting in a Receiver Operating Characteristic (ROC) curve. The critical signal and noise probability density function (pdf) assumptions from polyspectra theory which must be met for more effective noise suppression relative to classical second order power spectral methods are: (1) discussed in relation to detection results as reported in the literature review; and, (2) illustrated via examples using both direct and indirect nonparametric Discrete Fourier Transform (DFT) bispectrums employing Fast Fourier Transforms (FFT) for both simulated and real data. The signal set consisted of pure tones and a hop code used in active sonar. Signal detection, Bispectrum, Underwater acoustics, Gaussian noise, Higher order spectra, Polyspectra.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1994
- Accession Number
- ADA275227
Entities
People
- G. L. Morella
Organizations
- Pennsylvania State University