Abstract
Currently, WSN (Wireless Sensor Network) is the most standard services employed in commercial and industrial applications, because of its technical development in a processor, communication, and low-power usage of embedded computing devices. The WSN is built with nodes that are used to observe the surroundings like temperature, humidity, pressure, position, vibration, sound etc. The clustering is an efficient technique used to achieve the specific performance requirements of large scale wireless sensor networks. We have carried out the performance analysis of cluster-based wireless sensor networks for different communication patterns. In this work we are using fuzzy clustering; k-mean and self organizing map (SOM) based clustering method. It is observed that overhead in cluster based protocol is not much dependent upon update time. Simulation a result indicates that a cluster based protocol has low communication overheads compared with the velocity based protocol