Underdetermined direction-of-arrival (DOA) estimation for quasi-stationary signals impinging on a uniform circular array (UCA) with M sensors is addressed in this paper. We apply the Khatri-Rao (KR) approach to the UCA and obtain a new signal model which is capable of providing O(M2) sensors. Meanwhile, the virtual steering matrix can be decomposed into a product of characteristic matrix depending on the array structure and a Vandermonde matrix bearing the DOA information. The exact number of virtual sensors that the KR-UCA model is able to provide is studied as well. Simulation results are included to demonstrate the effectiveness of the proposed method.
- Khatri-Rao subspace approach
- Manifold separation technique
- Quasi-stationary signal
- Underdetermined estimation
- Uniform circular array