Cluster Head Selection and Packet Delivery Estimation Based on K-Means and FCM

Abstract

In this paper comparative study have been presented for the efficient cluster head selection based on k-means and fuzzy c-means (FCM) clustering algorithms. It is observed that the nodes assignment after the clustering is different through k-means and FCM. It is because of the variant initialization mechanism of the k-means and FCM. But the assignment of cluster does not affect the results. It is clearly depicted from the packet delivery time results by our approach. It shows that the k-means and FCM have the capability of CHs selection in the required time frame and it shows the effectiveness in different iterations also. When aggregate packet delivery has been considered the same situation has been observed which depicts the capability of our approach. K-means found to be faster in comparison to FCM.