A closed-form wideband direction-of-arrival estimation with chirplet-based adaptive signal decomposition algorithm

Abstract

The Direction-Of-Arrival (DOA) estimation with Coherent Signal Subspace (CSS) is not easy to extend from narrowband to wideband case. Time-frequency analysis is a powerful technique to deal with time-variant or non-stationary signal. Its combination with CSS explores a new field in signal processing, especially the wideband DOA estimation. The chirp function is one of the most fundamental functions in nature. Many nature events can be modeled as chirplet function, such as radar system or seismic exploring. Hence, the chirplet-based signal decomposition has attracted great attention in the area of signal processing. In this paper, the closed-form wideband DOA estimation with algorithm of chirplet-based adaptive signal decomposition is developed. With only three testing values, the delay of signals arriving at different elements can be acquired through a closed-form solution, then the DOA of sources can be got. Compared with Maximum Likelihood (ML), this novel approach has very low complexity. Its robust performance is demonstrated with numerical simulations.