A Tuned classification approach for efficient heterogeneous fault diagnosis in IoT-enabled WSN applications. (October 2021)

• Record Type: Journal Article
• Authors: Lavanya, S.; Prasanth, A.; Jayachitra, S.; Shenbagarajan, A.
• Is Part Of: Measurement. Volume 183(2021)
• Publication Date: 2021-10
• Subjects: Fault diagnosis -- Heterogeneous Faults -- Internet of Things -- Network Stability -- Wireless Sensor Network; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2021.109771
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: The novel methodology to classify the heterogeneous faults in IWSN. Three new diagnosis algorithms to detect multiple heterogeneous faults in IWSN. The simulation has been assessed in two different scenarios: IWSN#1 and IWSN#2. A new metric (Network Stability) is hosted to analyze the performance of IWSN. Abstract: The advancement of the Internet of Things (IoT) technologies will play a significant role in the growth of smart cities and industrial applications. Wireless Sensor Network (WSN) is one of the emerging technology utilized for sensing and data transferring processes in IoT-based applications. However, heterogeneous faults like hardware, software, and time-based faults are the major determinants that affect the network stability of IoT based WSN (IWSN) model. In this paper, a novel Energy-Efficient Heterogeneous Fault Management scheme has been proposed to manage these heterogeneous faults in IWSN. Efficient heterogeneous fault detection in the proposed scheme can be achieved by using three novel diagnosis algorithms. The new Tuned Support Vector Machine classifier facilitates to classify the heterogeneous faults where the tuning parameters of the proposed classifier will be optimized through Hierarchy based Grasshopper Optimization Algorithm. Finally, the performance results evident that the diagnosis accuracy of the proposed scheme acquires 99% and the false alarm rate sustains below 1.5% during a higher fault probability rate. The diagnosis accuracy rate is enhanced up to 17% as compared with existing techniques.