Cluster Formation Algorithm in WSNs to Optimize the Energy Consumption Using Self-Organizing Map

Wireless Sensor Networks (WSNs) are considered as one of the most prevailing technologies in today’s world due to the diversified applications. These applications are huge in the range such as environmental monitoring, health care, civil and military, disaster management to other surveillance systems. Minimization of energy is one of the most exciting tasks in WSNs as small sensor nodes are battery powered and deploy in remote environments. Clustering is one such imperative technique that can conserve energy more broadly, and evenly which makes the network operational for a longer period. This research paper aims at developing an energy-aware cluster-based routing protocol using Artificial Neural Network (ANN) which finds an optimal number of clusters and rotates the cluster head periodically to balance the energy consumption throughout the network. Typically, the proposed algorithm is developed based on Self-Organizing Map (SOM) to form the clusters, and the K-means algorithm is used to form different sizes of clusters. Finally, an optimal number of clusters is found that impacts the network load and balances the energy consumption. MATLAB is used as a simulation tool for experimental analysis. Simulation results prove that the proposed cluster-based routing protocol “LEACH-SOM” dominates the traditional LEACH routing protocol in terms of minimal energy consumption and makes the network active for long period.