<p>Accurate localization of unmanned aerial vehicle is crucial in global positioning system-denied environments. The existing localization-related approaches rely on the global positioning system signals, which are unreliable in certain conditions. The usage of inertial sensors in the localization methods is prone to sensor drift over time because that creates difficulties during accurate position estimation. To address these challenges, a novel acoustic localization method is proposed in this research article using extended Kalman filter. The array of acoustic sensors placed in the global positioning system-denied environment helps to detect time difference of arrival of signals for the acoustic localization. The proposed localization model is implemented for operating the unmanned aerial vehicle even in the limited global positioning system availability and reducing the effects of environmental noises and sensor drift. The advantages to integrate the extended Kalman filter model into time difference of arrival model are robust tracking capabilities and prediction ability of future position of unmanned aerial vehicle. For the experimental purpose, the real-time simulations are performed on the robot operating system, Gazebo and MATLAB environments. The robustness of the proposed method is evaluated based on the comparative analysis using the efficient performance measures of standard deviation and root-mean-square error. The proposed model provides significant potential for the applications of reliable unmanned aerial vehicle localization in the poor global positioning system coverage area and complex indoor environments. This research contributes more reliable and efficient localization with the potential implications for various applications such as disaster response, agriculture and security.</p>