TIME-MUSIC DOA estimation based on the exploitation of an arbitrary-order temporal structure in the data

This paper presents a unifying method called TIME-MUSIC for high-resolution direction of arrival estimation of statistically independent complex narrowband signals in noise with a planar sensor array of arbitrary geometry. It is based on the exploitation of an arbitrary-order temporal structure in the data by cumulants. A MUSIC-like spatial pseudo-spectrum is computed that exhibits peaks at the source directions of arrival. The theory is unifying in several senses. Firstly, it is general with respect to the order of the exploited temporal structure. Secondly, all types of statistical variability in the data, such as arbitrary-order nonstationarity and n on whiteness, are incorporated into the problem formulation in a unified manner. Finally, the conjugation pattern of the arguments of the involved cumulant functions can be chosen arbitrarily. In practice, this should be done in accordance with the characteristics of the involved signals. Depending on the number of sensors, the order of the exploited temporal structure, and the chosen conjugation pattern, a large number of source directions can be determined that exceeds the number of sensors for orders larger than one.