Unsupervised Machine Learning Based Group Head Selection and Data Collection Technique

The Detection Devices (DD) is used for various use cases like monitoring the environment, security data packets sending towards the control center. The sustainability of Detection Device Network (DDN) can be done by reducing the amount of energy dissipation along with increase in the hole- betterment ratio (HBR). When the DDs are used into multiple groups with each group having a group head DD which can communicate with other group. The existing methods make selection of group head either randomly or based on energy level or it is based on distance. This selection process involves monitoring of the network at regular intervals by other specific DDs. In this paper the selection of group head DD is performed by using energy, distance and packet delivery ratio using a K means-based method which can determine which will be better choice for group head DD based on closeness factor. The communication process for the control center is done collecting the data from main group DD, initiator DD and base station. The proposed method will be compared with LEACH based data collection and M-LEACH based data collection with respect to delay, hops, energy, and hole-betterment-ratio.