Distributed optical fiber anomaly detection based on IoT

Distributed optical fiber sensing systems play an important role in monitoring critical infrastructure such as pipelines and bridges, and can collect and analyze large amounts of data in real time. However, these systems face challenges when dealing with large amounts of data, especially in the areas of anomaly detection and resource allocation. The space-based Internet of things (Siot) makes up for the lack of terrestrial network coverage through satellite networks, provides global signal transmission over long distances and is suitable for data transmission and resource management in distributed optical fiber sensing systems. In this paper, an anomaly detection method based on LSTM-VAE is proposed, the integration of VAE and LSTM in the spatiotemporal data processing of distributed Internet of things system improves the accuracy and robustness of anomaly detection significantly. To further improve the system performance, context-aware multi-armed bandit (MAB) algorithm is introduced to dynamically assign tasks to iot devices. The MAB algorithm considers real-time context information, such as device status and network conditions, intelligently selects the most suitable device to handle specific tasks, and realizes efficient cooperation and resource optimization among multi-devices, it provides a new solution for distributed optical fiber sensor monitoring, and shows excellent performance in efficient data processing and resource allocation. Keywords: Distributed