Design and Evaluation of Real-Time Data Storage and Signal Processing in a Long-Range Distributed Acoustic Sensing...

Area of Science:

• Geophysics
• Data Science
• Cloud Computing

Background:

• Managing vast datasets from Distributed Acoustic Sensing (DAS) poses significant storage and processing challenges.
• Efficient cloud-based solutions are needed for real-time analysis of long-range DAS data.

Purpose of the Study:

• To develop and evaluate an efficient cloud-based data management pipeline for DAS.
• To assess the performance of cloud computing systems for DAS data processing using the CloudSim framework.

Main Methods:

• Implemented a pipeline system to efficiently transfer large DAS data volumes to Amazon Web Services (AWS) DynamoDB.
• Utilized the CloudSim framework to evaluate the performance of various virtual machine (VM) configurations for DAS data computations.

Main Results:

• The DynamoDB pipeline achieved low latency (40 ms per batch) and demonstrated scalability for DAS data storage.
• CloudSim analysis revealed that VM performance, processing elements, and MIPS significantly impact processing time, with optimal resource allocation crucial for efficiency.
• Increased fiber length and data volume showed improved processing efficiency, suitable for real-time monitoring.

Conclusions:

• The proposed DynamoDB-based pipeline offers a scalable and low-latency solution for cloud-based DAS data management.
• Cloud computing resources can be effectively optimized for DAS data processing, ensuring efficient real-time monitoring and feature extraction.
• Further research into resource optimization can enhance the performance and cost-effectiveness of cloud-based DAS systems.