An IDFT-based root-MUSIC for arbitrary arrays

Root-MUSIC algorithm, designed for uniform linear arrays, has been extended to arrays of arbitrary geometry by means of manifold separation techniques but at the cost of increased computational complexity. In this paper, an inverse discrete Fourier transform (IDFT)-based method is proposed in which polynomial rooting is avoided. The proposed method asymptotically exhibits the same performance as the extended root-MUSIC, implying that it outperforms the conventional MUSIC in terms of resolution ability. A remarkable property of this algorithm is that it has a computationally efficient implementation because a finite number of IDFT operations can run in parallel.