Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation

In this work, we propose a fault detection strategy for wireless sensor networks (WSNs) called Trend Correlation based Fault Detection strategy (TCFD). This strategy can detect the faulty sensor nodes through analyzing the trend correlation and the median value of neighboring nodes. On this basis, a...

Full description

Bibliographic Details
Main Authors: Xiuwen Fu, Ye Wang, Wenfeng Li, Yongsheng Yang, Octavian Postolache
Format: Article
Language:English
Published: IEEE 2021-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/9316657/
id doaj-8ac9c8d7a20a48ec9e1f9875966901f4
record_format Article
spelling doaj-8ac9c8d7a20a48ec9e1f9875966901f42021-03-30T15:29:57ZengIEEEIEEE Access2169-35362021-01-0199073908310.1109/ACCESS.2021.30498379316657Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend CorrelationXiuwen Fu0https://orcid.org/0000-0002-4405-7573Ye Wang1Wenfeng Li2Yongsheng Yang3Octavian Postolache4Institute of Logistics Science and Engineering, Shanghai Maritime University, Shanghai, ChinaInstitute of Logistics Science and Engineering, Shanghai Maritime University, Shanghai, ChinaSchool of Logistics Engineering, Wuhan University of Technology, Wuhan, ChinaInstitute of Logistics Science and Engineering, Shanghai Maritime University, Shanghai, ChinaISCTE-Lisbon University Institute, Lisbon, PortugalIn this work, we propose a fault detection strategy for wireless sensor networks (WSNs) called Trend Correlation based Fault Detection strategy (TCFD). This strategy can detect the faulty sensor nodes through analyzing the trend correlation and the median value of neighboring nodes. On this basis, aiming to avoid the excessive routing overhead caused by over-frequent fault detection, a fault detection self-starting mechanism is designed based on the cubic exponential smoothing method. Since the detection results in TCFD are determined by historical data at consecutive times and do not rely on the comparison of instantaneous sensed values at a single moment, it can significantly reduce the impact of fault detection time on detection accuracy. The simulation results have indicated that compared with referenced strategies, the proposed TCFD can obtain better fault detection accuracy for four common fault types of sensor nodes; in the case where the real fault rate of the network reaches 0.5, at least 70% of the faulty nodes can be detected by TCFD and the false alarm rate can still be kept below 30%; with the help of fault detection self-starting mechanism, the response time of sensor nodes to faults can be significantly shortened.https://ieeexplore.ieee.org/document/9316657/Wireless sensor networksfault detectiontrend correlationself-starting mechanismfault typeexponential smoothing method
collection DOAJ
language English
format Article
sources DOAJ
author Xiuwen Fu
Ye Wang
Wenfeng Li
Yongsheng Yang
Octavian Postolache
spellingShingle Xiuwen Fu
Ye Wang
Wenfeng Li
Yongsheng Yang
Octavian Postolache
Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation
IEEE Access
Wireless sensor networks
fault detection
trend correlation
self-starting mechanism
fault type
exponential smoothing method
author_facet Xiuwen Fu
Ye Wang
Wenfeng Li
Yongsheng Yang
Octavian Postolache
author_sort Xiuwen Fu
title Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation
title_short Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation
title_full Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation
title_fullStr Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation
title_full_unstemmed Lightweight Fault Detection Strategy for Wireless Sensor Networks Based on Trend Correlation
title_sort lightweight fault detection strategy for wireless sensor networks based on trend correlation
publisher IEEE
series IEEE Access
issn 2169-3536
publishDate 2021-01-01
description In this work, we propose a fault detection strategy for wireless sensor networks (WSNs) called Trend Correlation based Fault Detection strategy (TCFD). This strategy can detect the faulty sensor nodes through analyzing the trend correlation and the median value of neighboring nodes. On this basis, aiming to avoid the excessive routing overhead caused by over-frequent fault detection, a fault detection self-starting mechanism is designed based on the cubic exponential smoothing method. Since the detection results in TCFD are determined by historical data at consecutive times and do not rely on the comparison of instantaneous sensed values at a single moment, it can significantly reduce the impact of fault detection time on detection accuracy. The simulation results have indicated that compared with referenced strategies, the proposed TCFD can obtain better fault detection accuracy for four common fault types of sensor nodes; in the case where the real fault rate of the network reaches 0.5, at least 70% of the faulty nodes can be detected by TCFD and the false alarm rate can still be kept below 30%; with the help of fault detection self-starting mechanism, the response time of sensor nodes to faults can be significantly shortened.
topic Wireless sensor networks
fault detection
trend correlation
self-starting mechanism
fault type
exponential smoothing method
url https://ieeexplore.ieee.org/document/9316657/
work_keys_str_mv AT xiuwenfu lightweightfaultdetectionstrategyforwirelesssensornetworksbasedontrendcorrelation
AT yewang lightweightfaultdetectionstrategyforwirelesssensornetworksbasedontrendcorrelation
AT wenfengli lightweightfaultdetectionstrategyforwirelesssensornetworksbasedontrendcorrelation
AT yongshengyang lightweightfaultdetectionstrategyforwirelesssensornetworksbasedontrendcorrelation
AT octavianpostolache lightweightfaultdetectionstrategyforwirelesssensornetworksbasedontrendcorrelation
_version_ 1724179428619059200