Source analysis and separation using spatial filtering and time frequency distribution in array signal processing under complicated environments

Abstract

The paper proposes a method to analyze the sources. Under complicated environments, for example, there are different types of sources existing in the far field or more than one source in the same direction, common array processing methods do not consider these situations. But they will bring difficulty when we want to acquire the characteristics of these sources. Some sources may be indistinguishable in space, but they may be distinguishable in time-frequency domain, which we can use while analyzing the sources. The paper adopts spatial filtering and time frequency distribution (TFD) to deal with these situations. Spatial filtering can keep the sources in a particular direction from being interrupted by sources in other directions and extract the original signals in that direction, while TFD is used to analyze the extracted signals in time-frequency domain to get some information of signals of the sources, such as the frequency range. Simulation results prove the effectiveness of the method.