Derivative DFT Beamspace ESPRIT: a high performance closed-form 2-D arrival angle estimation algorithm

This paper presents Derivative DFT Beamspace ESPRIT, a new closed-form algorithm for direction-of-arrival (DOA) estimation with uniform linear arrays or uniform rectangular arrays. The algorithm uses a novel virtual derivative DFT beamforming procedure to improve upon the performance of the recently developed DFT Beamspace ESPRIT algorithm. This beamforming procedure yields an additional invariance relationship which the algorithm exploits to obtain higher estimation accuracy (the algorithm is shown to outperform both DFT Beamspace ESPRIT and Unitary ESPRIT). Further, Derivative DFT Beamspace ESPRIT possesses all the attractive features (such as low computational complexity, and the ability to provide automatically paired source azimuth and elevation angle estimates) of the two aforementioned algorithms.

[1]  Björn E. Ottersten,et al.  Multiple invariance ESPRIT , 1992, IEEE Trans. Signal Process..

[2]  Michael D. Zoltowski,et al.  Closed-form 3D angle estimation with rectangular arrays via DFT Beamspace ESPRIT , 1994, Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers.

[3]  Cherian P. Mathews Improved closed-form DOA/frequency estimation via ESPRIT using DFT and derivative DFT beamforming , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[4]  Sören Anderson On optimal dimension reduction for sensor array signal processing , 1993, Signal Process..