K-Means-Based Method for Clustering and Validating Wireless Sensor Network

This work is considered a clustering problem in a wireless sensor network, where sensor nodes are artificially generated and randomly distributed over the range of the network. The sensor is small in size, a short distance in communication, limited in storage space, and un-rechargeable battery. The task of this sensor is to sense the data from the area of being deployed. The clustering technique is employed to partition the area of the application into sub-areas; the distance-based method is used to partition the sensors in WSN. In this paper, we propose K-means-based method for clustering and validating grouping of sensor nodes by using external indices named purity. The proposed method can solve the clustering problem in WSN by partitioning the provided artificial sensor set into sub-clusters and validate them. The simulation result shows the ability of the proposed method in solving the power consumption by dividing the region of sensing and confirms that this method is suitable for large-scale wireless networks.