Study of Energy Efficient Clustering Algorithms
for Wireless Sensor Network

Abstract

Energy utilization and network life time are key issues in design of routing protocols for Wireless sensor network. Many algorithms have been proposed for reducing energy consumption and to increase network life time of the WSN. Clustering algorithms have gained popularity in this field, because of their approach in cluster head selection and data aggregation. LEACH (distributed) is the first clustering routing protocol which is proven to be better compared to other such algorithms. TL-LEACH is one of the descendants of LEACH that saves better the energy consumption by building a two-level hierarchy. It uses random rotation of local cluster base stations to better distribute the energy load among the sensors in the network especially when the density of network is higher. As the clusters are adaptive in LEACH and TL-LEACH, poor clustering set-up during a round will affect overall performance. However, using a central control scheme for cluster set-up may produce better clusters by distributing the cluster head nodes throughout the network. LEACH-C is another modification to LEACH that realizes the above idea and provides better results through uniform distribution of cluster heads avoiding redundant creation of cluster heads in a small area. In our project, we propose a centralized multilevel scheme called CML-LEACH for energy efficient clustering that assumes random distribution of sensor nodes which are not mobile. The proposed scheme merges the idea of multilevel hierarchy, with that of the central control algorithm providing uniform distribution of cluster heads throughout the network, better distribution of load among the sensors and improved packet aggregation. This scheme reduces energy consumption and prolongs network life time significantly as compared to LEACH, TL-LEACH and LEACH-C. The simulation results show comparisons of our scheme with the existing LEACH, TL-LEACH and LEACH-C protocols against chosen performance metrics, using Omnet++.