A signal enhanced state space approach for the estimation of two-dimensional frequencies

In this paper, we present a new algorithm for the two-dimensional (2-D) frequency estimation problem. The new method begins with the construction of a state space model associated with the noiseless 2-D data. Two auxiliary matrices are then introduced from which the two frequency components can be estimated via frequency shifting properties. Moreover, two signal enhancement procedures have also been incorporated to make the algorithm more robust. The resulting signal enhanced state space (SESS) algorithm not only can handle the case when some 2-D frequencies share one common component, but it also yields low deviation, high resolution frequency estimates. Simulation results justify this new algorithm