Improved DFT method for DOA estimation with extended coprime array:based on large difference coarray

In this paper, Direction of arrival estimation (DOA) in coprime array is investigated, and an advanced method for DOA estimation is presented with extended coprime array (ECA). Specifically, we use Discrete Fourier Transform (DFT) to obtain initial DOA estimation with large difference co-arrays generated by ECA, then perform Taylor series approximation and total least squares (TLS) method around the initial estimation to compensate the estimation offset. Compared to conventional DFT method, the proposed method enjoys better estimation accuracy with fewer computational burden since it avoids extra searches and benefits more from large difference co-array. Besides, in contrast to spatial smoothing based methods which halve array aperture, the proposed method avoids the process of spatial smoothing and fully utilises uniform degrees of freedom (DOFs). In this case, the proposed method has advantages on estimation accuracy and computational efficiency with multiple impinging signals, especially under the condition of large difference co-array. Furthermore, the proposed method is applicable to other types of advanced coprime array. The superiority of the proposed method has been proved by numerical simulations.