Performance of successive cancellation techniques for time of arrival estimation

Two successive cancellation methods for estimation of time or angle of arrival were presented by Krasny and Koorapaty (1986) and their performance was investigated for multipath channels with uncorrelated paths components. In this paper, the performance of the methods is investigated over channel models that have been designed to simulate real-world environments and have been extensively used for evaluating cellular positioning methods in standardization bodies. The first method performs successive cancellation with a correlation approach used for each iteration. The second method uses the multiple signal identification and classification (MUSIC) approach for each iteration. It is shown that the method based on the correlation approach outperforms MUSIC and Capon's (1969) minimum variance approach for many cases of interest. The performance for the methods are investigated in the context of time of arrival estimation.