Polynomial Manifold Interpolation for Direction Finding in the Presence of Mutual Coupling

Direction-of-arrival (DOA) estimation in the presence of mutual coupling among the antennas is an important problem in direction finding systems. To address the mutual coupling issue, a polynomial manifold interpolation approach is proposed in this paper. In this approach, a set of sample manifold vectors corresponding to pre-determined DOA grids are acquired and stored in advance in the direction finding receiver. The manifold vector corresponding to the DOA of interest can be obtained via polynomial interpolation over the sample manifold vectors. Then, MUtiple SIgnal Classification (MUSIC) algorithm can be used over the interpolated manifold vector to obtain the desired DOA estimation. Compared to the linear manifold interpolation, the density of the pre-determined DOA grids can be greatly reduced for polynomial manifold interpolation, and thus the burden for acquisition and storage of the sample manifold vectors can be also reduced substantially. Simulation results are also presented in this paper to demonstrate the effectiveness of the proposed approach.