Abstract With the significant increase in railway speeds, the potential hazards posed by external intrusions have become increasingly severe, presenting major risks to operational and public safety. Traditional monitoring systems, such as manual inspection and video surveillance, are inefficient and costly for continuous monitoring. In contrast, Distributed Acoustic Sensing (DAS) provides continuous monitoring capability, high sensitivity, and cost-effectiveness, making it highly valuable for railway perimeter safety monitoring. Therefore, we propose a two-dimensional (2D) localization method based on DAS, which, to the best of our knowledge, is the first to integrate spectral subtraction with the Multiple Signal Classification (MUSIC) algorithm for railway intrusion localization using DAS. Experimental results demonstrate that the method enables precise 2D localization of intrusion events within a range of approximately 10 meters, along the central axis of the fiber sensor arrays, and for frequencies within a certain range that satisfy narrowband and far-field conditions. It provides valuable support for comprehensive railway management and decision-making, with significant potential for broader application in railway safety.
Zhang et al. (Wed,) studied this question.