AN INTEGRAL OPERATOR BASED ADAPTIVE SIGNAL SEPARATION APPROACH

The operator-based signal separation approach uses an adaptive operator to separate a signal into additive subcomponents. And different types of operator can depict different properties of a signal. In this paper, we define a new kind of integral operator which can be derived from the second kind of Fredholm integral equation. Then, we analyze the properties of the proposed integral operator and discuss its relation to the second condition of Intrinsic Mode Function (IMF). To demonstrate the robustness and efficacy of the proposed operator, we incorporate it into the Null Space Pursuit algorithm to separate several multicomponent signals, including a real-life signal.