Optimized Cluster with Genetic Swarm Technique for Wireless Sensor Networks

Background/Objectives: In the current work, a Berkeley-Media Access Control (B-MAC) clustering protocol with a hybridized Genetic Algorithm (GA) as well as Particle Swarm Optimization (PSO) methods for overcoming the clustering issue through discovery of quantity of clusters, Cluster Heads as well as cluster members is proposed. Methods/Statistical Analysis: Wireless Sensor Networks (WSNs) are comprised of several quantities of minute nodes with restricted capabilities. The primary problem with these kinds of networks is the energy constraints. Plenty of research has been carried out in this field, with clustering emerging as the most efficient solution to the problem. The aim of clustering is the division of networks into groups with every group possessing a Cluster Head (CH). The job of Cluster Head is gathering, aggregating and transmitting data to Base Stations. Simulations using OPNET has been carried out in this study. Findings: The proposed protocol performance is tested for packet delivery ratio, end to end delay, number of hops to destination and jitter with various node mobility levels. The outcome reveals that the Local Search Binary PSO (LSBPSO) MAC Clustering performs better when compared with BMAC with flooding and BMAC with cluster based routing in either static or dynamic scenarios. Application/Improvements: Based on the performance of various MAC protocols, it is found that LSBPSO MAC clustering BMAC can be adopted for mobility based WSN applications like military recon operations, disaster management, security, healthcare systems, industrial mechanization and many others.