A Clustering Model for Collision Avoidance and Prevention of Packet Loss in a Biosensor Network

The packet loss in a biosensor network makes it difficult to continuously monitor and transfer patient information to medical centers. It lacks in complete analysis of the disease. This research article aims to propose a collision free biosensor network that prevents packet loss. We propose a K-Means clustering with the Euclidean distance and elbow method for optimal number of clusters creation in the biosensor network. The biosensors communicate with a cluster head, and the cluster head directs the messages of each cluster to the nano-router. When the information moves up the hierarchy, from the cluster head to the nano-router and then to the micro-gateway, redundant messages are eliminated. While the active biosensors interact with the cluster head, the remaining biosensors enter a low-power sleep mode to reduce the total energy consumption and boost the system's efficiency. We obtained an energy efficiency of 98% and a packet loss rate of 0.5%.