Energy Efficient Node Clustering in Sensor Networks

Nowadays devices with sensing ability have become a widely spread part of our life. They are useful for monitoring different phenomena and making processes of our life easier. These devices produce a large amount of data across wide sensor networks. Therefore, gathering and analysing data of sensor network has become an important issue. This process is challenging because sensor nodes that are part of these networks, have limited energy, computational and memory capacity. Limitation in energy especially affects sensor network lifetime. To face these challenges and to make this process more efficient, in this work it is proposed a new algorithm that introduces an energy efficient way for data acquisition in sensor networks. By using clustering techniques from data mining domain, it is suggested a self-organisation of sensor nodes into clusters. This is an important feature that makes this algorithm useful for most of sensor network applications, where nodes operate in an unsupervised way. Only a subset of nodes in the network have direct communication with distant central server. In this way, more energy is saved in communication and longer sensor network lifetime is provided.