This paper gives an overview of the popular Subspace based Direction of Arrival (DOA) estimation techniques. Mordern DOA estimation algorithms are based on subspace techniques. Subspace methods operate on a reduced parameter space by Eigen decomposition of the observed signal covariance matrix into two orthogonal spaces, the signal and noise subspaces and estimates the angles from one of these spaces. Subspace based DOA estimation techniques include; Multiple Signal Classification (MUSIC), Estimation of Signal Parameters via Rotational Invariance Technique (ESPRIT) and Unitary ESPRIT. These algorithms were originally formulated for single arrays. This paper extends their application to a bistatic MIMO (Multiple Input Multiple Output) radar system that have both a transmitting and a receiving array that are far apart. Performance comparison is made between the subspace based MUSIC algorithm and the spectral search based capon beam former using Matlab simulations. Several other Matlab simulations were run to demonstrate the application of ESPRIT and UNITARY ESPRIT algorithms on a bistatic MIMO System to estimate the Direction of Departure (DOD) and the Direction of Arrival (DOA) angles, the results showed how reduction in parameter space using subspace techniques reduces computational complexity and improves angle estimation resolution and accuracy.

Paper uploading is in process