Distributed self fault diagnosis algorithm for large scale wireless sensor networks using modified three sigma edit test

doi:10.1016/j.adhoc.2014.10.006

Ad Hoc Networks

Volume 25, Part A, February 2015, Pages 170-184

https://doi.org/10.1016/j.adhoc.2014.10.006 Get rights and content

Abstract

Distributed self diagnosis is an important problem in wireless sensor networks (WSNs) where each sensor node needs to learn its own fault status. The classical methods for fault finding using mean, median, majority voting and hypothetical test based approaches are not suitable for large scale WSNs due to large deviation in inaccurate data transmission by different faulty sensor nodes. In this paper, a modified three sigma edit test based self fault diagnosis algorithm is proposed which diagnose both hard and soft faulty sensor nodes. The proposed distribute self fault diagnosis (DSFD) algorithm is simulated in NS3 and the performances are compared with the existing distributed fault detection algorithms. The simulation results show that the detection accuracy, false alarm rate and false positive rate performance of the DSFD algorithm is much better in adverse environment where the traditional methods fails to detect the fault. The other parameters such as detection latency, energy consumption and the network life time are also determined.

Introduction

Wireless sensor networks (WSNs) comprise of thousands of tiny sensor nodes having limited memory, less processing power, battery constraint is deployed in an environment such as underwater, underground for multimedia and terrestrial applications like navigation, environmental monitoring, agriculture, landslide monitoring and emergency rescue operations [1], [43], [33]. The sensor nodes are sensing and processing capability and interact with other sensor nodes using a wireless medium. The sensor nodes form an adhoc network without following any infrastructure. In fact, the communications between sensor nodes is a type of peer to peer communication.

Since the sensor nodes are deployed in hostile and human inaccessible environments, they are subjected to various kinds of faults. These faulty sensor nodes lead to produce erroneous results during their normal operation. In order to prevent the WSNs from the effect of faulty sensor node, a self fault diagnosis algorithm is proposed here to diagnose both hard [3], [7], [13] and soft faulty sensor nodes [6], [17], [42], [2]. The fault free sensor nodes produce correct results during the network operation. The hard faulty sensor nodes do not respond, whereas the soft faulty sensor nodes respond with erroneous data skewed from the original data. The actual and erroneous data sensed by different sensor nodes is assumed to be a random variable which varies from one region to another region of the sensor network [18], [34]. Distributed self fault diagnosis of WSNs has recently important due to their application in various sectors of society. The dependence on WSN encompasses to design and develop a robust fault diagnosis algorithm for WSNs, so that they can sustain for longer duration in the presence of faults.

The fault diagnosis techniques based on classical estimates like sample mean, variance, co-variance or correlations are adversely influenced by large deviation of data for a faulty sensor node [25], [8], [30]. In fact, these estimators are producing correct fault status when many sensor nodes are faulty within a particular region. Motivated by this, a modified three sigma edit test approach is adapted to diagnose the faulty sensor nodes present in WSNs. In the proposed approach, the performance of the diagnosis depends on neighboring node’s data where each sensor node participates in the fault diagnosis process to identify itself as faulty (hard or soft) or fault free. The accuracy in finding the status of all the sensor nodes depends on the number of neighboring nodes. The proposed diagnosis algorithm performs better when more number of neighboring nodes are likely to be faulty.

It has been seen from the literature that the existing method leads to a large number of message exchanges over the network for data and status exchange which rapidly depletes the energy of the sensor nodes. It really puts a large overhead for the large scale WSNs. Due to poor performance and high energy overhead of the existing set of approaches, it is necessary to design and develop an efficient fault diagnosis algorithm for large scale WSNs.

This paper has following contributions: (i) Modified three sigma edit test based fault diagnosis algorithm is proposed to diagnose the faulty sensor nodes present in WSNs. (ii) The proposed method is compared with traditional comparison model and three sigma edit test. (iii) A distributed self-fault diagnosis (DSFD) algorithm where each sensor node diagnoses itself with high detection accuracy, low false alarm rate and false positive rate. (iv) Evaluation of the DSFD algorithm using standard simulator NS3 and compare the performance with the existing works in the literature given by Panda and Khilar [29], Chen et al. [6] and Jiang [17].

The remaining part of the paper is organized as follows. In Section 2, the related work which provides an exhaustive survey about the previous work is discussed. The network, fault and radio model are discussed in Section 3. The proposed DSFD algorithm is described in Section 4. The analysis of the algorithm along with modified three sigma edit test method is discussed in Section 5. The simulation results with discussions are given in Section 6. Finally, Section 7 concludes the paper.

Section snippets

Related works

The system and node level fault diagnosis has been traditionally known as a system level diagnosis after the work of the researchers Preparata et al. in [32]. Since then, all the system level diagnosis algorithms are feasible for multi-computer and multi-processor systems. A generalized theory for system level fault diagnosis is presented in [39]. The necessary and sufficient conditions are provided for any fault-pattern of any size to be uniquely diagnosable, under the symmetric, and

System model

The system model consists of network, fault and radio model. In network model, the network topology and the way nodes communicate with each other are specified. In fault model, we describe different types of fault based on the behavior of faulty, fault free sensor nodes, and the data generated by different sensor node. Along with this, these models also describe the behavior of the sensor nodes on the occurrence of faults. Finally, the radio model is developed to calculate the energy required

Distributed self-fault diagnosis algorithm (DSFD)

The DSFD algorithm is consisting of two phases such as initialization and self-diagnosis phase. In initialization phase, each sensor node $s_{i}$ transmits a message containing the sensed data $x_{i}$ to its neighboring nodes ${Neg}_{i}$ and waits for a estimated transmission time ${ETT}_{i}$ as given in (18). During that transmission time, it also collects all the messages coming from its neighbors. After ${ETT}_{i}$ expires, each node $s_{i}$ extracts the information from all receiving messages and maintains a neighboring Table

Analysis of the DSFD Algorithm

The proposed DSFD algorithm for WSNs is based on the assumption of network and fault model given in Section 3. Let the data of the sensor node $s_{i}$ at kth time instant is denoted as $x_{i} (k)$ . In order to find the faulty sensor node in the network, the data ${x_{i} (k)}_{i = 1}^{N}$ of all the sensor nodes are to be analyzed. The sensor reading $x_{i} (k)$ can be either actual sensed data or the erroneous data. This $x_{i} (k)$ follows normal distribution $N (A, σ_{i}^{2})$ where $σ_{i}^{2}$ is the variance of erroneous data present at $i$ th

Simulation results and discussions

The performance of the proposed DSFD algorithm is evaluated and compared with the existing DFD [29], Jiang [17] and JSA [6] algorithms in terms of detection accuracy (DA), false positive rate (FPR), false alarm rate (FAR), and other network parameters such as total number of message exchange in the network, energy consumption, diagnosis latency to find all the faulty node, network life time which are defined in Section 5. All the algorithms are simulated in a standard network simulator NS3 [37]

Conclusion

In this paper, a modified three sigma edit test based distributed self fault diagnosis (DSFD) algorithm for large scale wireless sensor networks is proposed. Here each sensor node collects data from the neighbors and then diagnose itself by using the modified three sigma edit test of the received data. The performance of the DSFD is compared with the existing algorithms and all the algorithms are simulated in NS3 simulator. The simulation results show that the proposed method outperforms over

Meenakshi Panda received her M-Tech degree in CSE from Utkal University, India in 2006. Currently she is pursuing her Ph.D in the department of CSE at National Institute of Technology Rourkela, India. She worked as Asst. Prof. at Siliocon Institute of Technology (SIT) Bhubaneswar, Orissa, India for four year before joining Ph.D. Her research interest includes fault detection in wireless sensor networks and robust distributed estimation in faulty WSN.

References (43)

I.F. Akyildiz et al.
Wireless sensor networks: a survey
Sci. Direct Trans. Comput. Netw.
(2002)
Indrajit Banerjee et al.
Effective fault detection and routing scheme for wireless sensor networks
Comput. Electr. Eng.
(2014)
D.D. Geeta et al.
Fault tolerance in wireless sensor network using hand-off and dynamic power adjustment approach
J. Netw. Comput. Appl.
(2013)
Bill C.P. Lau et al.
Probabilistic fault detector for wireless sensor network
Expert Syst. Appl.
(2014)
M.H. Lee et al.
Fault detection of wireless sensor networks
Sci. Direct J. Comput. Commun.
(2008)
T. Pham-Gia et al.
The mean and median absolute deviations
Math. Comput. Model.
(2001)
Maneesha Vinodini Ramesh
Design, development, and deployment of a wireless sensor network for detection of landslides
Ad Hoc Netw.
(2014)
Xiaolong Xu et al.
LEDFD: a low energy consumption distributed fault detection algorithm for wireless sensor networks
Int. J. Distrib. Sens. Netw.
(2014)
Jennifer Yick et al.
Wireless sensor network survey
Comput. Netw.
(2008)
Prabir Barooah et al.
Cut detection in wireless sensor networks
IEEE Trans. Parall. Distrib. Syst.
(2012)

Ronald P. Bianchini et al.

Implementation of online distributed system-level diagnosis theory

IEEE Trans. Comput.

(1992)

David Braun, Sarit Mukherjee, Cuneyt Akinlar, Zero Configuration Networking, US Patent 7,002,924, February...

Jinran Chen et al.

Distributed fault detection of wireless sensor networks

S. Chessa et al.

Crash faults identification in wireless sensor networks

Sci. Direct J. Comput. Commun.

(2002)

M. Ding, D. Chen, K. Xing, X. Cheng, Localized fault-tolerant event boundary detection in sensor networks, in: INFOCOM...

E.P. Duarte et al.

Distributed diagnosis of dynamic events in partitionable arbitrary topology networks

IEEE Trans. Parall. Distrib. Syst.

(2012)

Elias Procópio Duarte et al.

A hierarchical adaptive distributed system-level diagnosis algorithm

IEEE Trans. Comput.

(1998)

Mourad Elhadef, Azzedine Boukerche, Hisham Elkadiki, Diagnosing mobile ad-hoc networks: two distributed...

Harald T. Friis, A note on a simple transmission formula, in: Proceedings of the I.R.E. and Waves and Electrons, May...

W.R. Heinzelman, A. Chandrakasan, H. Balakrishnan, Energy-efficient communication protocol for wireless microsensor...

S.H. Hosseini et al.

On self-fault diagnosis of the distributed systems

IEEE Trans. Comput.

(1988)

Cited by (102)

A fault localization approach based on multi-system PCA and dynamic SDG: Application in train lifting equipment
2024, Robotics and Computer-Integrated Manufacturing
The development of the current rail transportation is increasingly rapid, and train maintenance is closely related to the high-efficiency running of rail transportation. Train lifting equipment (TLE) is important for train maintenance, for which stable operation is crucial. TLE fault is characterized by early abnormally weak and capable of propagation. Early anomaly detection and fault diagnosis are the keys to TLE fault maintenance, and this paper proposes a TLE fault localization methodology to address these two problems. For early anomaly detection, a combination of sequence-based and principal component analysis-based methods is utilized to address the problem. For fault diagnosis, the signed directed graph-based method is utilized to ensure the completeness of reasoning results, and the system division mechanism and dynamic division mechanism are utilized to narrow down the scope of faults and accelerate the reasoning process, respectively. The experiments prove that the proposed method can reduce the loss caused by equipment faults and improve the reliability and stability of the equipment.
Comprehensive fault diagnosis in UAV-assisted sensor networks: A three-phase automated approach
2024, Computer Communications
Ensuring operational safety in an automated working environment is a crucial requirement. In this article, we propose a technique for automated fault diagnosis in unmanned aerial vehicle-based sensor networks (UAV-sensor networks) aimed explicitly at the composite nature of different faults. Our technique consists of 3-stages: (1) UAVs collecting data from sensors and running a modified Z-score statistics to diagnose faulty readings; (2) the UAV cluster head running a multivariate analysis of variance (MANOVA) to identify data faults in cluster members; and (3) utilizing a probabilistic neural network with correlation centroids at the ground control station for further diagnosis of collected UAV data. This approach is intended for progressive improvement in fault diagnosis at different stages of data collection in a UAV-sensor network for critical applications. We conduct an extensive evaluation of our 3-stage approach on the data collected in real-time from the sensors in a physically created UAV-sensor network. The performance of the proposed modified Z-score test improves fault detection accuracy of $\sim$ 20.8%, false alarm rate of $\sim$ 54.3%, and false positive rate of $\sim$ 67.2% with the traditional and existing sensor fault detection approach. The performance of MANOVA to identify the permanent UAV data fault shows fault detection accuracy of $\sim$ 10.9%, false alarm rate of $\sim$ 44.3%, and false positive rate of $\sim$ 48.7% superior to the intermittent and transient faults. Our preliminary results suggest that the proposed approach can handle multiple composite faults and outperforms existing research results.
Self-detection based fault diagnosis for wireless sensor networks
2023, Ad Hoc Networks
Wireless sensor network (WSN) is a collection of low-power and low-cost sensor nodes (SNs) deployed stochastically in the monitoring area to support various applications. The random occurrence of faulty sensor nodes affects the quality of service in WSN. Most of the existing fault detection algorithms are based on statistics, thresholds, majority voting, hypothetical tests, comparison, or machine learning and depend on the neighbor's sensing information to detect the fault status of SNs. Hence, these fault detection algorithms suffer from low detection accuracy (DA) and a high false alarm rate (FAR). Moreover, existing fault detection techniques are not scalable for large-scale WSNs, as they suffer from high energy overhead and detection latency. Furthermore, existing fault diagnosis algorithms burden an SN's resources while estimating the fault status of the WSN. This paper proposes a deep belief network (DBN) based self-detection algorithm to address these issues. The performance of the proposed DBN based self-detection algorithm has been evaluated using google colab, and MATLAB. The simulation results of the proposed algorithm has been compared with the existing fault detection algorithms. Simulation results show that the proposed algorithm performs better than the existing algorithms in terms of DA, FAR, and false positive rate. The performance of the proposed algorithm has also been evaluated in other network parameters such as energy consumption, and fault detection latency.
Cloud-fog-based approach for smart wildfire monitoring
2023, Simulation Modelling Practice and Theory
Wildfires have disturbed the delicate balance of the Earth's ecosystem, exacerbating global warming, floods, droughts, and causing a reduction in oxygen level. Early detection of wildfires is crucial in order to mitigate the risk of losses. The Internet of Things (IoT) has emerged as a critical tool in bridging disparate technologies, allowing for the real-time collection of data from sensors and further analysis of factors influencing the occurrence of wildfires (FIWs) on fog devices or cloud servers. However, most deployment of these sensor nodes in hostile and unmanned environments presents significant challenges to the reliability information collection of FIWs and the effectiveness of energy-conserving technologies. To address these challenges, this paper proposes a novel wildfire monitoring model based on an IoT-Fog-Cloud three-tier architecture. This model incorporates energy-conserving and fault-tolerant scheduling algorithms in the fog layer, ensuring that sensor nodes operate reliably. In the cloud layer, the multi-Grained Cascade Forest (GC-Forest) algorithm is utilized to analyze both spectral and FIWs data. Furthermore, in order to accurately assess the spread trend of forest fires, the Wang Zhengfei model is improved. Experimental results demonstrate that the energy-conserving algorithm reduces energy consumption by approximately 25.48%, while the fault-tolerant scheduling algorithm remains a fault detection rate of over 98.7%. The GC-Forest algorithm exhibits a low error rate of only 16.25%, with a hit rate of 83.75%, and a Hanssen-Kuiper skill score of 63.73%.
Fault Detection and Diagnosis Methods for Sensors Systems: a Scientific Literature Review
2023, IFAC-PapersOnLine
Innovation technology in industrial devices allows enhancing process monitoring and control and have a significant impact on communication and interfacing systems within all production contexts, both in the process industries and in the manufacturing sector. Asset status and performances can be monitored by traditional sensors and intelligent devices, wired and wireless, able to collect data and processing them or make them available to specific processing algorithms. Reliability and maintenance researchers and practitioners are focusing on models, tools, and techniques for in field and real-time detection of process anomalies and equipment fault detection. Most analyses start from the assumption that sensors and monitoring systems guarantee reliability and operational availability and that they are always able to provide timely and correct data. Therefore, malfunctions of such devices could affect trust on data-driven analysis and related decision-making process on asset management and workers’ safety. In recent years, different approaches were explored, aimed at recognizing failure mechanisms in industrial measuring devices and sensors systems, both using traditional methodologies or implementing intelligent algorithms, to accurately prevent and predict the anomalies and identify their nature and position before they generate malfunctioning or interruption of plant's operations. This paper proposes a systematic analysis of the scientific literature related to fault/failure detection and diagnosis in sensors and monitoring systems, to obtain an updated state-of-the-art and identify the most promising approaches and research challenges on this topic. A particular focus is dedicated to the usability of soft sensors and artificial intelligence algorithms, in data-driven models.
Leak detection and localization in water distribution networks using conditional deep convolutional generative adversarial networks
2023, Water Research
This paper explores the use of ‘conditional convolutional generative adversarial networks’ (CDCGAN) for image-based leak detection and localization (LD&L) in water distribution networks (WDNs). The method employs pressure measurements and is based on four pillars: (1) hydraulic model-based generation of leak-free training data by taking into account the demand uncertainty, (2) conversion of hydraulic model input demand-output pressure pairs into images using kriging interpolation, (3) training of a CDCGAN model for image-to-image translation, and (4) using the structural similarity (SSIM) index for LD&L. SSIM, computed over the entire pressure distribution image is used for leak detection, and a local estimate of SSIM is employed for leak localization. The CDCGAN model employed in this paper is based on the pix2pix architecture. The effectiveness of the proposed methodology is demonstrated on leakage datasets under various scenarios. Results show that the method has an accuracy of approximately 70% for real-time leak detection. The proposed method is well-suited for real-time applications due to the low computational cost of CDCGAN predictions compared to WDN hydraulic models, is robust in presence of uncertainty due to the nature of generative adversarial networks, and scales well to large and variable-sized monitoring data due to the use of an image-based approach.

View all citing articles on Scopus

P.M. Khilar was born in India, on July 10, 1968. Graduated from Mysore University in Computer Science and Engineering in the year 1990. Got his M.Tech Degree from NIT, Rourkela in the year 1999 and received his Ph.D. from IIT, Kharagpur in the year 2009. Presently, he is working as an Assistant Professor in CSE Department at NIT Rourkela, India. His current research interests includes Parallel and Distributed Processing and Fault Tolerant Computing.

View full text

Distributed self fault diagnosis algorithm for large scale wireless sensor networks using modified three sigma edit test

Abstract

Introduction

Section snippets

Related works

System model

Distributed self-fault diagnosis algorithm (DSFD)

Analysis of the DSFD Algorithm

Simulation results and discussions

Conclusion

Sci. Direct Trans. Comput. Netw.

Comput. Electr. Eng.

J. Netw. Comput. Appl.

Expert Syst. Appl.

Sci. Direct J. Comput. Commun.

Math. Comput. Model.

Ad Hoc Netw.

Int. J. Distrib. Sens. Netw.

Comput. Netw.

Cut detection in wireless sensor networks

IEEE Trans. Parall. Distrib. Syst.

Implementation of online distributed system-level diagnosis theory

IEEE Trans. Comput.

Distributed fault detection of wireless sensor networks

Crash faults identification in wireless sensor networks

Sci. Direct J. Comput. Commun.

Distributed diagnosis of dynamic events in partitionable arbitrary topology networks

IEEE Trans. Parall. Distrib. Syst.

A hierarchical adaptive distributed system-level diagnosis algorithm

IEEE Trans. Comput.

On self-fault diagnosis of the distributed systems

IEEE Trans. Comput.