Abstract:
Prolonging the lifetime of a sensor node powered by the battery in homogeneous wireless sensor
networks (WSNs) is a very popular research area now-a-days. Many research works including low
energy adaptive clustering hierarchy (LEACH) and its variants use clustering techniques to reduce
energy spent by keeping most of the nodes in the sleeping mode whenever possible. These schemes
however suffer from a number of ways. Firstly, they fail to provide any mathematical model for
estimating the expected lifetime of the WSN that are influenced by the initial energy and the
number of sensor nodes as well as the number of clusters. Lack of this measurement puts
limitations on making the optimum usage of energy. Secondly, the round time estimated in LEACH
and its variants are fixed and cannot adapt with the changing scenarios of elapsing the energy and
death of cluster heads. Finally, the existing schemes estimate the cluster head probability of each
node either not considering the remaining energy or reducing the number of clusters, which
ultimately results in energy wastage and information loss. In this thesis, a new energy probability
based deterministic life span LEACH approach has been proposed by addressing all of the
abovementioned problems. A mathematical model is proposed for estimating the lifetime of WSN
based on numbers of nodes and clusters. This model can therefore be used for calculating the
requirement of initial energy and the optimum number of cluster heads to achieve an expected
lifetime. The dynamic round time model proffered in the proposed scheme reduces the energy
wastage through electing the sensor nodes as cluster heads more rapidly as the energy of sensor
node reduces. The proposed scheme also uses the remaining energy to calculate the probability of
sensor nodes so that the low energy nodes, having possibility of dying prematurely, are refrained
from being cluster heads and hence increases energy utilization while reducing packet dropping
rates. The superiority of the proposed scheme, studied through extensive simulations, has therefore
setup new dimensions in WSN technology and hence hastening the arrival of using modem sensing
and monitoring services by deploying sensor nodes.
