Detection and identification of cyclostationary signals

Abstract

Propeller noise can be modeled as an amplitude modulated (AM) signal. Cyclic Spectral Analysis has been used successfully to detect the presence of analog and digitally modulated signals in communication systems. It can also identify the type of modulation. Programs for Signal Processing based on compiled languages such as FORTRAN or C are not user friendly, and MATLAB based programs have become the de facto language and tools for signal processing engineers worldwide. This thesis describes the implementation in MATLAB of two fast methods of computing the Spectral Correlation Density (SCD) Function estimate, the FFT Accumulation Method (FAM) and the Strip Spectral Correlation Algorithm (SSCA), to perform Cyclic Analysis. Both methods are based on the Fast Fourier Transform (FFT) algorithm. The results are presented and areas of possible enhancement for propeller noise detection and identification are discussed.