Delay-Based Dynamic Clustering Method for the IoT Cluster
Seo Jin Chang, Boyeon Kim, Yunseok Chang · 2022
A dynamic clustering method can reconstruct IoT nodes into logical group units when using several groups of IoT nodes. However, massive control at a logical group level does not guarantee optimal control performance. To efficiently deliver and execute a massive control command within a certain communication delay, it is necessary to cluster each logical group into subgroups and execute massive control commands by each subgroup. This paper proposes the DBDC(Delay-Based Dynamic Clustering) method, a clustering method based on DBSCAN that can optimize the communication delay of a massive control within a logical group. Through several simulations, the parameters of the DBSCAN algorithm create different clustering cases, and the DBDC method can find the optimal parameter value that satisfies the optimization condition. Therefore, this study shows that the DBDC method can effectively design a cluster system so collective IoT clusters have the best massive control performance.