Energy efficient cluster-head election scheme for wireless sensor network based on fuzzy load balancing

• Main Author: Nancy, anak Bundan
• Format: Thesis
• Language: English
• Published: Universiti Malaysia Sarawak, (UNIMAS) 2014
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://ir.unimas.my/id/eprint/8346/1/Nancy%20Bundan%20ft.pdf; http://ir.unimas.my/id/eprint/8346/
• Institution: Universiti Malaysia Sarawak

The fundamental component of Wireless Sensor Networks (WSNs) is sensor nodes. As sensor nodes are generally battery-powered, the lifetime of WSNs were greatly influenced by the fact that the main source of energy consumption is communication. Thus, once deployed, the entire network lifetime shows a strong dependency on the battery lifetime of individual nodes and WSNs can only subsist on while the battery power is adequate. One of the popular solutions in overcoming this handicap is by using the clustering technique wherein the sensor nodes were divided into several groups, by which one of the sensor nodes will be appointed as cluster-head within each group. The cluster-head is responsible to aggregate data collected by its cluster member and transmit it back to the base station. In this research, a centralized cluster-head election scheme using fuzzy logic was proposed since fuzzy logic is capable of blending different types of parameters and imitating the situations from the real-world more closely. The „Inverse Betweenness Centrality Load-Balancing fuzzy clustering scheme (IBL)‟ uses node‟s energy, node‟s centrality and distance as the fuzzy descriptor to elect cluster-head. To justify the efficiency of the proposed IBL fuzzy clustering scheme, comparisons have been made with the schemes proposed by other researcher via computer simulations. Simulation results demonstrate that the proposed IBL fuzzy clustering scheme performs better in terms of network lifetime, the number of alive nodes and cluster formation showing overall percentage increase of 7% to 26% for 50 nodes and 2% to 18% for 100 nodes.