Impacts of wireless sensor networks strategies and topologies on prognostics and health management

Abstract

In this article, we used Wireless Sensor Network (WSN) techniques for monitoring an area under consideration, in order to diagnose its state in real time. What differentiates this type of network from the traditional computer ones is that it is composed by a large number of sensor nodes having very limited and almost nonrenewable energy. A key issue in designing such networks is energy conservation because once a sensor depletes its resources, it will be dropped from the network. This will lead to coverage hole and incomplete data arriving to the sink. Therefore, preserving the energy held by the nodes so that the network keeps running for as long as possible is a very important concern. If we achieve to improve the network lifetime and Quality of Service (QoS). Diagnosing the state of area will be more accurate for a longer time. One of the most important elements to achieve a QoS in WSN is the network coverage which is usually interpreted as how well the network can observe a given area. Obviously, if the coverage decreases over time, the diagnosis quality decreases accordingly. Various coverage strategies are thus proposed by the WSN community, in order to guarantee a certain coverage rate as long as possible, to reach a certain QoS that in turn will impact the diagnosis and prognostic quality. Various other strategies are in common use in WSN like data aggregation and scheduling, to preserve a QoS in wireless sensor networks, as long as possible. We argue that such strategies are not neutral if this network is used for prognostic and health management. Some politics may have a positive impact while other ones may blur the sensed data, like data aggregation or redundancy suppression, leading to erroneous diagnostics and/or prognostics. In this work, we will show and measure the impact of each WSN strategy on the resulting estimation of diagnostics. We emphasized several issues and studied various parameters related to these strategies that have a very important impact on the network, and therefore on data diagnostics over time. To reach this goal, to evaluate both prognostic and health management with the WSN strategies, we have used six diagnostic algorithms.

Coverage Prediction for Target Coverage in WSN Using Machine Learning Approaches

“…The reinforcement learning based approaches are based on the application of Q-learning to solve the different issues in WSN such as coverage, routing and QoS parameters. The machine learning based approaches provide these advantages in the solution of coverage and connectivity problems [29]:…”

Research and Diagnostics for the Laboratory of Pressure Resistant Sensors

“…The spring continues to push constant pressure into the sensor. When the combination of these two types of creating power is possible to diagnose, test and monitor various processes [Dao 2015, Farhat 2019, Chudzikiewicz 2011].…”

An Integrative Framework for Online Prognostic and Health Management Using Internet of Things and Convolutional Neural Network

“…Internet of Things and wireless sensor networks (WSNs) are widely used to get data in the industry [21]. Data aggregation is the strategy of wireless sensor networks [22], which is used in PHM for the communication of information between the equipment. Internet of Things is a network system that connects equipment with sensors, hardware network, and cloud servers [10,23].…”