Use this URL to cite or link to this record in EThOS:
Title: Energy efficient clustering algorithms for wireless sensor networks
Author: Zanjireh, Morteza Mohammadi
ISNI:       0000 0004 5923 6608
Awarding Body: Glasgow Caledonian University
Current Institution: Glasgow Caledonian University
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Wireless Sensor Networks (WSNs) have had remarkable advances in the past couple of decades due to the flexibility of their applications and their successful deployment in many fields. In order to supervise an area, hundreds or thousands of sensors can be placed in the environment and can collaborate with each other. The resulting aggregated data can then be delivered to a base station. The adaptable and distributed nature of WSNs has made them popular in a broad range of applications. Clustering, a method in which nodes are organised into groups with a Cluster Head (CH), is a widely accepted approach to the organisation of sensor nodes to address concerns about network congestion and energy efficiency. In this approach, the number and distribution of the OHs are both fundamental to the energy efficiency and flexibility of the clustering method. Some networks' individual sensors may vary in their importance, owing to their varying proximity to critical regions. This situation requires its own clustering method. Hence in this thesis, two new distributed energy efficient clustering algorithms for VvSNs and their analytical models have been proposed for use with those networks that have sensors of equal 01' varying importance. For those networks with sensors of equal importance, the Avoid Near Cluster Heads (ANCH) algorithm is proposed which reduces the network's energy consumption and prolongs its lifetime significantly. This is achieved by optimising the distribution of OHs across the network. An analytical model for the ANCH algorithm is presented in order to predict the energy consumption of the network. The results of the extensive simulation study show considerable reduction in network energy consumption which prolonged the network lifetime on average by 119.13%, 59.40%, and 17.16% compared with those of the Low Energy Adaptive Clustering Hierarchy (LEACH), Hybrid Energy-Efficient Distributed (HEED), and Low Energy Adaptive Clustering Hierarchy with Sliding Window and Dynamic ~urnber of Nodes (LEACH-S\VDN) algorithms, respectively. In addition, the proposed analytical model was about 96.56% accurate in predicting the energy consumption under different operating conditions. The proposed analytical model showed that the energy consumption pattern of ANCH is not influenced by the number of clusters, or the closeness factor, which is defined as the proximity of CHs. On the other hand, for WSNs with sensors of varying importance, the Activity-aware Avoid Near Cluster Heads (A-ANCH) algorithm is proposed which prioritises the prolonging of active sensors by way of monitoring the important regions in a network environment and extends the lifetimes of active sensors by sacrificing inactive ones. An analytical model for the AANCH algorithm is presented in order to predict the number of sensed events (packets) in the network. The results of the extensive study for this type of environment show a considerable increase in the number of sensed events on average by 85.58%, 67.40%, and 46.61%, compru'ed with those of LEACH, HEED, and ANCH algorithms, respectively. rvIoreover, the experiments illustrated on average 91.50% accuracy for the proposed mathematical model compared with the simulation results. The analytical model presented here has also revealed that the number of sensed events in the A-ANCH algorithm is sensitive to the number of sensors, the proportion of active sensors to all sensors and the network dimensions. On the other hand, it is almost insensitive to the number of clusters.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available