Mixed near field and far field sources localization algorithm based on MEMS vector hydrophone array

• Published in: Measurement : journal of the International Measurement Confederation Vol. 151; p. 107109
• Main Authors: Shang, Zhenzhen, Zhang, Wendong, Zhang, Guojun, Zhang, Xiaoyong, Ji, Songxiang, Wang, Renxin
• Format: Journal Article
• Language: English
• Published: London Elsevier Ltd 01.02.2020; Elsevier Science Ltd
• Subjects: Algorithms; Arrays; Computer simulation; Direction of arrival; Estimating techniques; Fresnel zones; Hydrophones; Localization; Magnetic fields; Mathematical analysis; Matrix; Matrix algebra; Matrix methods; MEMS vector hydrophone; Mixed sources; MUSIC algorithm; Port and starboard ambiguity; Steering; Vector space; Mixed sources; MUSIC algorithm; Port and starboard ambiguity; MEMS vector hydrophone
• ISSN: 0263-2241; 1873-412X
• DOI: 10.1016/j.measurement.2019.107109

•We propose a mixed fields algorithm for MEMS vector hydrophone array.•We analyze the characteristics of steering vector matrix.•The direction vector array is decomposed into different matrices.•The DOA and range are separated efficiently.•Simulation test have proven the advantage of MEMS vector hydrophone array. A mixed near and far field sources localization algorithm is proposed which applied to the uniform linear MEMS vector hydrophone array. The algorithm includes three steps: firstly, the steering vector contains common azimuthal information in both near and far field, direction-of-arrival (DOA) can be obtained by using the MUSIC algorithm. Secondly, the direction vector array is decomposed into different matrices and DOA and range are separated efficiently. All DOAs can be obtained through searching for a feature matrix. Thirdly, one-dimensional spectrum search in Fresnel zone is carried out to realize the accurate estimation of range in near-field. Finally, the performance of the algorithm was simulated. The results verify that the proposed algorithm can achieve DOA estimation within the range of [0:2*π] and solve the problem of port and starboard ambiguity. The vector matrix algorithm does not require construction the high-order cumulants and multi-dimensional search, and can avoid parameter matching procedure.