State Space-Based Method for the DOA Estimation by the Forward-Backward Data Matrix Using Small Snapshots

Abstract

In this presentation, a new low computational burden method for the direction of arrival (DOA) estimation from noisy signal using small snapshots is presented. The approach introduces State Space-based Method (SSM) to represent the received array signal, and uses small snapshots directly to form the Hankel data matrix. Those Hankel data matrices are then utilized to construct forward-backward data matrix that is used to estimate the state space model parameters from which the DOA of the incident signals can be extracted. In contrast to existing methods, such as MUSIC, Root-MUSIC that use the covariance data matrix to estimate the DOA and the sparse representation (SR) based DOA which is obtained by solving the sparsest representation of the snapshots, the SSM algorithm employs forward-backward data matrix formed only using small snapshots and doesn't need additional spatial smoothing method to process coherent signals. Three numerical experiments are employed to compare the performance among the SSM, Root-MUSIC and SR-based method as well as Cramér–Rao bound (CRB). The simulation results demonstrate that when a small number of snapshots, even a single one, are used, the SSM always performs better than the other two method no matter under the circumstance of uncorrelated or correlated signal. The simulation results also show that the computational burden is reduced significantly and the number of antenna elements is saved greatly.