LAUSR

Direction estimation in the presence of the Doppler effect can be carried out according to the maximum likelihood principle, which requires solving a nonlinear multidimensional problem. A complex multidimensional problem can be transformed into simple one-dimensional problems through signal separations. In the existing localization method for noncircular sources that takes into account the Doppler effect based on signal separation, computationally expensive eigen-decomposition of an NxN matrix is required at each search point where N is the number of snapshots. This paper presents an efficient method that requires no eigen-decomposition. The proposed method finds direction estimates by optimization of the same cost function of two variables that the existing method handles. Their estimation performances are thus identical, although the complexity of the former is far less than that of the latter. The optimization of the cost function with respect to one variable is theoretically solved without eigen-decomposition, which leads to lower complexity.