Files in this item



application/pdfCREDC_tianyuan_klara.pdf (237kB)
(no description provided)PDF


Title:Resilient Data Collection in Refinery Sensor Networks Under Large Scale Failures
Author(s):Liu, Tianyuan; Guo, Hongpeng; Lui, King-Shan; Jin, Haiming; Nahrstedt, Klara
Subject(s):data collection
network resilience
self-healing protocol
oil & gas refinery
Abstract:Wireless sensors and measurement devices are widely deployed in oil and gas refineries to monitor the health of the pipes. These sensors are deployed along the pipes in an open area and thus are subject to large scale failures due to cyber-physical attacks and hazardous environments. In this paper, we study the resilience issues in collecting data from a dense and large scale set of sensors deployed over the physical refinery pipe network. We construct a multi-tree sensor mesh network over the refinery sensors for data collection. The reporting messages within one of the trees, while passing along the tree, are protected by a secret key shared among all sensors on the tree. Our construction aims to minimize the data collection time and ensures that the information leakage probability of the secret key is bounded. To tolerate large scale failures, we present a distributed self-healing protocol, which enables a tree node to discover a secondary path when its parent fails. The simulation result shows that the self-healing protocol tolerates large scale failures with high probability and has small overhead in data collection time.
Issue Date:2017
Citation Info:Tianyuan Liu, Hongpeng Guo, King-Shan Lui, Haiming Jin, Klara Nahrstedt, "Resilient Data Collection in Refinery Sensor Networks Under Large Scale Failures"
Genre:Technical Report
Date Available in IDEALS:2017-09-20

This item appears in the following Collection(s)

Item Statistics