Intrusion Identification for Long-Distance Pipeline Based on Cloud-Edge Collaboration

DOFS based on phase-sensitive optical time-domain reflectometry (Φ-OTDR) are commonly used to detect and locate third-party events that could potentially damage long-distance pipelines. The proliferation of DOFS devices has led to an exponential growth in data generation at the edge of networks. To address the data intensive challenge and meet the requirements of low latency, this paper proposes a hierarchical cloud-edge col-laboration scheme for intrusion identification along long-distance pipelines. Computation-intensive tasks in the cloud and latency-sensitive operations on edge devices are performed in parallel, improving resource utilization rates. Deep learning algorithm is deployed at the edge of the computing network to identify the hazardous events. We evaluate event detection performance of four distinct events using a DOFS system deployed along the pipelines and experimental results demonstrate the feasibility.