Bearing estimation without using complex transform

A new eigen-decomposition method of direction of arrival (DOA) estimation for multiple narrowband sources is presented. The most important feature of the method hinges upon the fact that both the signal and eigenvector are real values and the principal components of the input stochastic process can be computed by neuron-like networks such as the APEX method. Thus, the method is simple in computation and beneficial to real-time processing. However, it may introduce uncertainty in its spectrum. The problem is resolved by using two uniform linear arrays. One is placed horizontally and the other is set at a slant. Simulation results show the merit of the new method for bearing estimation of incoherent sources in an additive white noise environment