Properties of Toeplitz approximation method (TAM) for direction finding problems

This paper gives the analysis of Toeplitz Approximation Method (TAM) for direction finding problem. There are essentially two steps in TAM. The first step is the apporximation procedure of a parameter set and the second step consists of extracting the information (angular direction and amplitude) from parameter set. We will show how the structure of a toeplitz matrix formed from a set of Markov parameters can be exploited to efficiently solve the approximation problems. By employing the Singular Value Decomposition (SVD) technique we obtain an approximation state-space realization. We will give many desirable features of our approach. These include a good numerical performance, stability of the realization, robustness and flexibility in determining model order.