Tehran University of Medical Sciences

Science Communicator Platform

Stay connected! Follow us on X network (Twitter):
Share this content! On (X network) By
Sound Source Localization Using Time Differences of Arrival; Euclidean Distance Matrices Based Approach Publisher



Zad Tehrani AK1 ; Makkiabadi B2 ; Pourmohammad A3 ; Hozhabr SH4
Authors
Show Affiliations
Authors Affiliations
  1. 1. Department of Electrical Engineering, Amirkabir University of Technology, Tehran, Iran
  2. 2. Department of Medical Physics and Biomedical Engineering, Tehran University of Medical Sciences, Tehran, Iran
  3. 3. Electrical Engineering Department, Amirkabir University of Technology, Tehran, Iran
  4. 4. Research Center for Biomedical Technologies and Robotics, Institute for Advanced Medical Technologies (IAMT), Tehran, Iran

Source: 9th International Symposium on Telecommunication: With Emphasis on Information and Communication Technology# IST 2018 Published:2018


Abstract

In the source localization problem, time differences of arrival (TDOA) and intensity level differences (ILD) of microphones can be employed to estimate the source location. Due to existing additive noise in real applications the ILD measurement provides less reliable information compared to the TDOA. Therefore, this study is focused on developing algorithms employing the TDOA information only. In the past studies, TDOA were used mostly for estimation of direction of the arrival (DOA) parameter. To find the source location from TDOA of different microphones, the intersection of several equations must be calculated which this solving process requires complex numerical analysis. The solving processes, which generally ignore the noise existence, are not robust to noise and might not converge to the true answer in real-world applications. This paper tackles the source localization problem by converting the numerical analysis approach to an iterative minimization one in order to improve localization accuracy in noisy conditions. The performance of the proposed iterative minimization algorithm is seen to be sensitive to the initial values. To address this issue, another algorithm, based on Euclidean Distance theory, is developed to obtain stable and accurate results. The proposed framework works properly in different SNR conditions. The results show that the proposed methods are more accurate than the existing numerical analysis based methods in different noisy conditions even in very low SNR conditions. © 2018 IEEE.