A Novel Data Gathering Approach Using Machine Learning Technique In Wireless Sensor Networks
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In recent years, intensive research is performed in the field of
newlineWireless Sensor Networks (WSN), especially for monitoring,
newlinecharacterizing, and tracking large physical environments and conditions
newlinesuch as, temperature, wind, pressure, and humidity. Due to its
newlineversatility, it has received significant attention and has been deployed in
newlineapplications such as natural disaster relief, military target tracking,
newlinehazardous environment exploration and wildlife monitoring. To process
newlinedata at remote central unit (e.g., sink), many of these application use
newlinesensors to periodically sense and send the sensory information. After
newlinedeployment, in many of these applications, the critical task is in
newlineclassifying the huge amount of sensed data and maintaining the sensors.
newlineTherefore energy efficient data collection protocol has become utmost
newlineimportance in these networks. With restricted energy storage capability
newlineof sensors and bandwidth scarcity, it is crucial to jointly consider
newlineguaranteed data delivery incorporated with efficient data gathering in
newlineWSN. Moreover, challenging deployment environments pose intricacies
newlinein reliable data transmission in these networks.
newlineThough many improvements are done, lifetime of the network
newlineremains a major issue in WSN technology. To develop a performancecentric
newlineautomated system, the work in this dissertation dedicatedly
newlineconcentrates on building a smart network management with reconfigurable
newlinesensors. It primarily focuses in achieving successful data
newlinegathering for critical applications using genetic algorithm and machine
newlinelearning techniques.
newlinevi
newlineThe first stage of the dissertation involves implementing a
newlinenovel quotProactive Event and Time driven (Pro_ET)quot protocol that
newlineexploits on-demand reconfigurable smart collectors for efficient energy
newlinemanagement using weighted fairness queuing (WFQ) mechanism. The
newlinerole of the Smart Collector (SC) is to self-organize by itself to exhibit
newlinethe capability to gather and aggregate data efficiently both during
newlinecritical and non-critical occasions. Implementing Pro_ET shows
newlinepromisin