Block Toeplitz with Toeplitz block covariance matrix for space-time adaptive processing

Interference covariance matrix estimations and computational complexity are two main concerns in space-time adaptive processing (STAP). This paper deals with the two problems by exploring a structured covariance matrix. First, a conjugate gradient iterative algorithm (CGIA) with reduced computational complexity is presented, which is based on FFT by using a block Toeplitz with Toeplitz block (BTTB) structure of the interference covariance matrix. To ensure the convergence of CGIA, an iterative BTTB (IBTTB) covariance matrix approximation is also proposed. Based on the approximated BTTB matrix, the corresponding STAP algorithm provides superior and robust performance both in limited sample support and in the presence of system errors.