An Energy Efficient Self Organizing Map Based Clustering Protocol For IoT Networks

The clustering technique is an optimal configuration for the Internet of Things (IoT) networks. It offers several benefits, such as energy conservation, latency reduction, and scalability. Meanwhile, energy consumption remains a major concern. In this regard, we introduce a new clustering approach based on the Self-Organizing Map algorithm (SOM), called Energy Efficient SOM (EESOM), conscious of energy consumption with an energy-aware cluster-head (CH) rotation policy that considers the current energy of cluster nodes and their distance from the winning neuron to determine the best CH. The dynamic CH-rotation avoids unbalanced energy consumption for the successive CHs in the cluster and reduces their premature death. When choosing a CH with the minimum distance to the winning neuron, the one with the minimum Euclidean distance to the member nodes will be elected. Consequently, the energy needed by members to send the collected data to their CH is reduced, and the lifetime of the network can be extended. Simulation results indicate that EESOM effectively reduces energy consumption and spreads the network lifespan.