Joint DOA and fundamental frequency estimation methods based on 2-D filtering
Jesper Rindom Jensen, Mads Græsbøll Christensen, Søren Holdt Jensen · VBN Forskningsportal (Aalborg Universitet) · 2010
It is well-known that filtering methods can be used for processing of signals in both time and space. This comprises, for example, fundamental frequency estimation and direction-of-arrival (DOA) estimation. In this paper, we propose two novel 2-D filtering methods for joint estimation of the fundamental frequency and the DOA of spatio-temporarily sampled periodic signals. The first and simplest method is based on the 2-D periodogram, whereas the second method is a generalization of the 2-D Capon method. In the experimental part, both qualitative and quantitative measurements show that the proposed methods are well-suited for solving the joint estimation problem. Furthermore, it is shown that the methods are able to resolve signals separated sufficiently in only one dimension. In the case of closely spaced sources, however, the 2-D Capon-based method shows the best performance.