Graph Kernel Based Clusterhead Selection Algorithm in Internet of Things

The Internet of Things (IoT) is easy to deploy, dynamic and applicable, and has attracted extensive attention from researchers. Clustering is one of the effective methods to measure the performance of IoT with different attributes. In this paper, we propose a Graph Kernel based Clusterhead selection Algorithm in IoT (GKCA). The key idea of GKCA algorithm is to find the clustering scheme in the process of cluster head selection mechanism. The performance of GKCA algorithm, such as average end-to-end delay, packet delivery ratio, was experimentally studied by using network simulation (NS2) software. Experiments show that the GKCA has a good performance advantage in dynamic IoT.