This Is AuburnElectronic Theses and Dissertations

Energy-efficient Clustering Method for Wireless Sensor Networks

Date

2010-07-14

Author

Lee, Kyeongdon

Type of Degree

thesis

Department

Computer Science

Abstract

Wireless sensor networks have recently emerged as an important computing platform. These sensors are power-limited and have limited computing resources. Therefore the sensor energy has to be managed wisely in order to maximize the lifetime of the network. There have been many studies on considering sensor's energy. Among these, clustering sensor nodes is a more efficient and adaptive approach in sensor networks having some essential requirements. First, every node in a cluster should elect one clusterhead. Second, clustering must minimize overhead. Third, network topology must be stable. Finally, as sensor nodes are operated; low power, low energy consumption is required. In this thesis, previous work on LEACH protocol is reviewed and a study on the issues related to the network lifetime is carried out.Simply speaking, LEACH requires the knowledge of energy for every node in the network topology used. LEACHs threshold which selects the clusterhead is fixed so this protocol does not consider network topology environments. Whereas other protocols, namely HEED and PEGASIS, which are hierarchical routing protocols, do not guarantee the number of clusterheads. Thus, ELP (Enhanced LEACH Protocol) algorithm is proposed, which selects clusterheads using di erent thresholds. New clusterhead election probability consists of the current energy of nodes that is compared to initial energy and current energy of neighbor nodes. In order to evaluate the energy efficiency for the ELP algorithm, the network lifetime of both LEACH and ELP is compared. The network lifetime of ELP shows an increase of more than 23% compared to the LEACH network lifetime. In conclusion, the proposed protocol efficiently manages the energy of sensor nodes resulting in an increase of the network lifetime.