Constructions of ϵ-mono-components and mathematical analysis on signal decomposition algorithm

The concept of mono-component is widely used in non-stationary signal processing and time-frequency analysis. A special class of mono-components, called ϵ-mono-components, were proposed in our recent publication. It was illustrated that this model coincides with the intuition of human beings on the physical mono-components very well provided that the parameter ϵ is sufficiently small. It is then very meaningful to construct desired ϵ-mono-components and design algorithms to decompose and represent non-stationary signals adaptively. This paper studies the constructions of ϵ-mono-components and makes mathematical analysis on an adaptive signal decomposition algorithm based on ϵ-mono-components.