Wavelet-based multiresolution analyses of signals
Mark Russell Kalmbach · Defense Technical Information Center (DTIC) · 1992
Signal analysts have traditionally relied on the Discrete Fourier Transform and various data windowing schemes for signal detection and classification. Some signals, notably those of a transient nature, are inherently difficult to analyze with these traditional tools. The Discrete Wavelet Transform has recently generated considerable interest in several areas of digital signal processing and a determination of its suitability as a signal analysis tool is necessary. Associated with wavelet theory is the concept of multiresolutional analyses which allow examination of a signal at different scales. This thesis investigates dyadic discrete wavelet decompositions of signals. A new multiphase wavelet transform is proposed and investigated. The multiphase transform technique is shown to be useful in transient signal analysis. Several MATLAB TM programs that perform multiresolutional analyses with various supporting features are provided.