Machine Learning Approach for Enhancing the Lifetime of the Wireless Sensor Network

Abstract

Wireless Sensor Networks (WSNs) play a crucial role in wireless data transmission. A WSN newlineconsists of a Base Station having many sensor nodes and these nodes are deployed randomly newlineacross the entire monitoring region. However, energy conservation is a major challenge in the newlineWSN as the long term usefulness of WSN mainly relies on the lifetime of the sensor nodes. newlineSince these nodes are made operational using a battery, their lifetime essentially depends on their newlinebattery source, whose replacement is not feasible. Over time, the nodes drain their energy in newlinesensing the region of interest. Thus, the only way to achieve the long lifetime of WSNs is newlinethrough the conservation of battery energy. This study examined the current WSN energy-saving newline newlinetechniques and suggested four different algorithms: (i) K-Means-PSO-GSO, (ii) K-Means-GSO- newlineKGMO, (iii) FCM-PSO-GSO, (iv) RSOM, (v) RSOM-WOEWMA with sleep active strategy and newline newlinewithout EWMA-based energy harvesting technique and (vi)RSOM-EWMA with sleep active newlinestrategy and EWMA-based energy harvesting technique. The research examined the performance newlineof the proposed algorithms by comparing them with the existing standard algorithms, e.g. newlineLEACH, PSO-PSO-WSN, EBC-S and NEEC. The performance metrics, such as Dead/Alive newlinenodes, number of packet sent transmitted, energy consumption and throughput, were used for newlinemeasuring the efficiency of the algorithms. It is observed that the proposed solutions clearly newlineoutperforms the existing algorithms. newline