An Efficient Energy Aware Clustering Protocol for WSN with Sink Mobility

number: 
3703
إنجليزية
Degree: 
Author: 
Mariam Rushdi Abd Al-Redha
Supervisor: 
Dr. Osama A. Awad
year: 
2016

Recent developments in wireless sensor network fields allow it to be suitable for sensing specified parameter(s) related to a certain environment. Routing the data wirelessly in energy efficient manner is the main task of network layer, thus, clustering routing scheme is one of the most efficient techniques for WSN for achieving this requirement. This thesis presents two centralized cluster based routing protocols for WSNs with sink mobility: Cluster Head selection based on Fuzzy Logic implemented on Zone Routing Protocol (CHFL-ZRP) & Mobile Cluster Head selection based on Fuzzy Logic implemented on Zone Routing Protocol (MCHFL-ZRP). The proposed protocols work for both static and dynamic sensor nodes; where CHFL-ZRP is applied in an environment that covers fixed nodes only, while MCHFL-ZRP can work with both fixed or/and mobile nodes. The selection of the cluster head nodes depends on applying fuzzy logic approach by incorporating three extracted node features, the centrality, the concentration and residual energy. A predictable sink mobility pattern for data gathering mechanism is applied by making the sink moves in a hexagonal path pattern and selecting the most appropriate diagonal size for the hexagon with regards to the network life time and the average energy consumption. Different simulation scenarios are considered in evaluating the performance using NS2.35-Software installed on Ubuntu-14.04 distribution, a Linux operating system which is virtualized on oracle VM-Ware workstation. The overall system is implemented on Windows-8.1 that packs a Core i3 CPU. Simulation are divided into two parts, the first one is performed on CHFL-ZRP with fixed and mobile sink. Different size of clusters and different diagonal path patterns for the sink movement are examined to exhibit their impact on the appearance of first node to die metric and on the consumed energy. The second part involved the evaluation and performance comparison of the developed CHFL-ZRP & MCHFL-ZRP protocols with the well-known LEACH, LEACH-C, LEACH-ERE, LEACH-ME, LEACH-M and CHEF protocols. Simulation results show that the CHFL-ZRP outperforms CHEF and LEACH-ERE in terms of half nodes to die and networks life time. Also, it outperforms the LEACH and LEACH-C in terms of the throughput, end to end delay, average energy consumption, networks life time, and half nodes to die. While the MCHFL-ZRP outperforms the LEACH-ME and LEACH-M in terms of the average remaining-consuming energy and average end to end delay.