Direction of arrival estimation in sensor arrays with faulty elements

Date

2020

Author

Külbay, Tayfun

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

27
views

6
downloads

Cite This

During the operational lifetime of a sensor array, certain sensors cease to operate properly. The identification of these sensors is an important problem. Furthermore, the operation of the array is continued even with the faulty sensors. Therefore, new techniques are required in order to ensure that the system estimation performance does not degrade significantly. In this thesis, two different fault types in a sensor array are considered. The first fault type is special to MEMS microphone arrays. MEMS microphones are usually constructed with pairwise sensors which share the common data channel in positive and negative cycles of a clock signal. When one of the MEMS microphones is faulty, same data is seen for both of the microphone channels and determining the faulty sensor becomes a major problem. A new technique is developed to determine the faulty microphone based on the use of a source with known direction-of-arrival (DOA). This approach can find the faulty sensor or sensors accurately in low SNR depending on the source position with respect to the array. In a second fault type, it is assumed that each sensor’s fault is associated with a probability. In other words, a sensor is either on or off with a probability “p” at any time. This problem is investigated in the literature before and G-MUSIC algorithm is proposed for a consistent DOA estimation. In this thesis, the problem scenario is investigated and a new approach is proposed to estimate the DOA with a faulty sensor array. It is shown that the performance of the proposed algorithm is better than G-MUSIC.

Subject Keywords

Estimation theory., DOA Estimation, missing data, faulty sensor, MEMS microphone

URI

http://etd.lib.metu.edu.tr/upload/12625344/index.pdf
https://hdl.handle.net/11511/45585

Collections

Graduate School of Natural and Applied Sciences, Thesis