Separation of mixed signals in dynamic environments: formulation and some implementation

We describe algorithms and some chip implementations for separating unknown, but independent, mixtures of temporal signals in static and dynamic environments. One algorithm is a modified HJ algorithm; the other is derived based on optimal control theory so as to maximize a functional which characterizes the linear independence of temporal signals. Moreover, we present some chip results and computer simulations and quantify the performance in static and dynamic environments.