Modified clustering algorithms and their performance for wireless sensor networks

Abstract

Wireless sensor networks (WSNs) are a type of Adhoc wireless networks. WSN consists of resource-constrained sensor nodes with battery as power source. These wireless nodes monitor the environment or an object and provide the collected data to central location. These networks are used in many newlineapplications, ranging from disaster warning systems to structural health newlinemonitoring of bridges for effective maintenance. Of late, research is underway newlineto study using WSNs for safety critical and non-safety critical systems of newlineaircrafts operations. WSNs are also useful in adding additional sensors compared to cabled system, some of which may be used as standby/redundancy sensors in aircraft monitoring systems. Replacing cabling with wireless sensors also helps to reduce the weight of aircraft, network complexity, cost of maintenance, and at the same time improves fuel efficiency. Despite the diversity of WSN applications, cost-performance ratios are yet not low. In WSNs, power saving is one of the most important research topics that helps improve cost-performance ratio. However, the variety of WSN deployment environments and application scenarios imply that there is no single solution to solve the energy issue and optimize network lifetime. newlineOne of the commonly used techniques to optimize energy consumption is clustering and forming cluster-trees. Several algorithms proposed earlier are newlinepresented in literature [2] to form clustered trees. In these algorithms, cluster heads (CH) are selected first, and then, the remaining nodes associate with one of the Cluster Heads (CHs) in order to form clusters. There are several approaches for electing CH. For instance, Low Energy Adaptive Clustering Hierarchy (LEACH) follows uniform probability approach to elect cluster heads. Although a few techniques have been developed in the past on the performance enhancement of LEACH protocol [1, 3, 8], none of them considered variable probability for cluster head selection and the same is proposed in D-LEACH for possible energy efficiency.